AW: Load Balancing and Session Timeouts
Connection Settings are only relevant for midtier -> ar-server connections. And indeed, there is no need to have any persistence on this type of connection. If this connection is also routed through a loadbalancer in front of your ar-servers, you can disable persistence there. But this will definitely not solve your problem. KR Conny Von: ARSList Im Auftrag von JD Hood Gesendet: Dienstag, 5. März 2019 17:19 An: ARSList Betreff: Re: Load Balancing and Session Timeouts If you are referencing the "Connection Settings" in the mid tier config, yes we have tried that (and it remains configured), but it did not help and the issue persists. The "Connection Lifespan" is set to 60 minutes and the "Connection Timeout" is 0 (which is infinite, correct?). Tomcat and Netscaler both are configured for 65 minute sessions. Question to all -- if the "Connection Settings" in the mid tier are configured, should we remove any session persistence settings form the Netscaler? Thanks, -JDHood On Tue, Mar 5, 2019 at 10:54 AM LJ LongWing mailto:lj.longw...@gmail.com>> wrote: JD, If you are on 9.x, you can configure Session Sharing (clustering) in your Mid-Tier containers, and then you wouldn't need sticky bits (persistence) On Tue, Mar 5, 2019 at 8:49 AM JD Hood mailto:hood...@gmail.com>> wrote: All, We have a simple load balancing arrangement between a Netscaler and two SSL Mid tiers (v18.08). End users are sporadically encountering session timeout errors when they would not be expected. Sometimes within a few minutes of connecting and in the midst of creating a new incident, they will get a session timeout error (arerr 9201) and they will have to refresh the page and re-authenticate. I've found the following BMC Support KB and shared it with the team: https://bmcsites.force.com/casemgmt/sc_KnowledgeArticle?sfdcid=kA21400d7ZtCAI Our netscaler resource advises that the cookie-based persistence option is not available when he is configuring load-balancing. I believe he is configuring persistence as IP based. Other than that, we can't identify any non-recommended configuration issues. I suspect that the cookie-based load balancer configuration is significant, but not being a LB expert, it's outside of my skillset and my most convincing argument at the moment is, "...because BMC Support recommends it..." Does anyone have any experience with a basic, simple Netscaler configuration to balance two SSL Mid tiers based on least number of connections (or any other strategy)? Or perhaps session timeouts are a known issue for other reasons that I have yet to stumble across? Thanks, -JDHood -- ARSList mailing list ARSList@arslist.org<mailto:ARSList@arslist.org> https://mailman.rrr.se/cgi/listinfo/arslist -- ARSList mailing list ARSList@arslist.org<mailto:ARSList@arslist.org> https://mailman.rrr.se/cgi/listinfo/arslist -- ARSList mailing list ARSList@arslist.org https://mailman.rrr.se/cgi/listinfo/arslist
Re: Load Balancing and Session Timeouts
If you are referencing the "Connection Settings" in the mid tier config, yes we have tried that (and it remains configured), but it did not help and the issue persists. The "Connection Lifespan" is set to 60 minutes and the "Connection Timeout" is 0 (which is infinite, correct?). Tomcat and Netscaler both are configured for 65 minute sessions. Question to all -- if the "Connection Settings" in the mid tier are configured, should we remove any session persistence settings form the Netscaler? Thanks, -JDHood On Tue, Mar 5, 2019 at 10:54 AM LJ LongWing wrote: > JD, > If you are on 9.x, you can configure Session Sharing (clustering) in your > Mid-Tier containers, and then you wouldn't need sticky bits (persistence) > > On Tue, Mar 5, 2019 at 8:49 AM JD Hood wrote: > >> All, >> >> We have a simple load balancing arrangement between a Netscaler and two >> SSL Mid tiers (v18.08). >> >> End users are sporadically encountering session timeout errors when they >> would not be expected. Sometimes within a few minutes of connecting and in >> the midst of creating a new incident, they will get a session timeout error >> (arerr 9201) and they will have to refresh the page and re-authenticate. >> >> I've found the following BMC Support KB and shared it with the team: >> >> https://bmcsites.force.com/casemgmt/sc_KnowledgeArticle?sfdcid=kA214000000d7ZtCAI >> >> Our netscaler resource advises that the cookie-based persistence option >> is not available when he is configuring load-balancing. I believe he is >> configuring persistence as IP based. Other than that, we can't identify any >> non-recommended configuration issues. >> >> I suspect that the cookie-based load balancer configuration is >> significant, but not being a LB expert, it's outside of my skillset and my >> most convincing argument at the moment is, "...because BMC Support >> recommends it..." >> >> Does anyone have any experience with a basic, simple Netscaler >> configuration to balance two SSL Mid tiers based on least number of >> connections (or any other strategy)? >> >> Or perhaps session timeouts are a known issue for other reasons that I >> have yet to stumble across? >> >> Thanks, >> -JDHood >> >> >> -- >> ARSList mailing list >> ARSList@arslist.org >> https://mailman.rrr.se/cgi/listinfo/arslist >> > -- > ARSList mailing list > ARSList@arslist.org > https://mailman.rrr.se/cgi/listinfo/arslist > -- ARSList mailing list ARSList@arslist.org https://mailman.rrr.se/cgi/listinfo/arslist
AW: Load Balancing and Session Timeouts
Hi, don’t know if it’s a good idea to have a setup like this. From my point of view session sharing is used to provide some kind of high availability. We setup session sharing and had a config mistake regarding too short timeouts. This led to frequent session “hopping” between several midtier instances. This in turn corrupted the session information, because more than one midtier was updating the session data within a short timeframe, overwriting some critical things. I highly recommend to use Cookie based persistence. In our case we do not rely on loadbalancer features. We’re using apache httpd in reverse proxy configuration to accomplish this. The loadbalancer routes to at least two httpd instances. Httpd takes care of routing the request to the right midtier based on session cookie. KR Conny Von: ARSList Im Auftrag von LJ LongWing Gesendet: Dienstag, 5. März 2019 16:54 An: ARSList Betreff: Re: Load Balancing and Session Timeouts JD, If you are on 9.x, you can configure Session Sharing (clustering) in your Mid-Tier containers, and then you wouldn't need sticky bits (persistence) On Tue, Mar 5, 2019 at 8:49 AM JD Hood mailto:hood...@gmail.com>> wrote: All, We have a simple load balancing arrangement between a Netscaler and two SSL Mid tiers (v18.08). End users are sporadically encountering session timeout errors when they would not be expected. Sometimes within a few minutes of connecting and in the midst of creating a new incident, they will get a session timeout error (arerr 9201) and they will have to refresh the page and re-authenticate. I've found the following BMC Support KB and shared it with the team: https://bmcsites.force.com/casemgmt/sc_KnowledgeArticle?sfdcid=kA21400d7ZtCAI Our netscaler resource advises that the cookie-based persistence option is not available when he is configuring load-balancing. I believe he is configuring persistence as IP based. Other than that, we can't identify any non-recommended configuration issues. I suspect that the cookie-based load balancer configuration is significant, but not being a LB expert, it's outside of my skillset and my most convincing argument at the moment is, "...because BMC Support recommends it..." Does anyone have any experience with a basic, simple Netscaler configuration to balance two SSL Mid tiers based on least number of connections (or any other strategy)? Or perhaps session timeouts are a known issue for other reasons that I have yet to stumble across? Thanks, -JDHood -- ARSList mailing list ARSList@arslist.org<mailto:ARSList@arslist.org> https://mailman.rrr.se/cgi/listinfo/arslist -- ARSList mailing list ARSList@arslist.org https://mailman.rrr.se/cgi/listinfo/arslist
Re: Load Balancing and Session Timeouts
JD, If you are on 9.x, you can configure Session Sharing (clustering) in your Mid-Tier containers, and then you wouldn't need sticky bits (persistence) On Tue, Mar 5, 2019 at 8:49 AM JD Hood wrote: > All, > > We have a simple load balancing arrangement between a Netscaler and two > SSL Mid tiers (v18.08). > > End users are sporadically encountering session timeout errors when they > would not be expected. Sometimes within a few minutes of connecting and in > the midst of creating a new incident, they will get a session timeout error > (arerr 9201) and they will have to refresh the page and re-authenticate. > > I've found the following BMC Support KB and shared it with the team: > > https://bmcsites.force.com/casemgmt/sc_KnowledgeArticle?sfdcid=kA21400d7ZtCAI > > Our netscaler resource advises that the cookie-based persistence option is > not available when he is configuring load-balancing. I believe he is > configuring persistence as IP based. Other than that, we can't identify any > non-recommended configuration issues. > > I suspect that the cookie-based load balancer configuration is > significant, but not being a LB expert, it's outside of my skillset and my > most convincing argument at the moment is, "...because BMC Support > recommends it..." > > Does anyone have any experience with a basic, simple Netscaler > configuration to balance two SSL Mid tiers based on least number of > connections (or any other strategy)? > > Or perhaps session timeouts are a known issue for other reasons that I > have yet to stumble across? > > Thanks, > -JDHood > > > -- > ARSList mailing list > ARSList@arslist.org > https://mailman.rrr.se/cgi/listinfo/arslist > -- ARSList mailing list ARSList@arslist.org https://mailman.rrr.se/cgi/listinfo/arslist
Load Balancing and Session Timeouts
All, We have a simple load balancing arrangement between a Netscaler and two SSL Mid tiers (v18.08). End users are sporadically encountering session timeout errors when they would not be expected. Sometimes within a few minutes of connecting and in the midst of creating a new incident, they will get a session timeout error (arerr 9201) and they will have to refresh the page and re-authenticate. I've found the following BMC Support KB and shared it with the team: https://bmcsites.force.com/casemgmt/sc_KnowledgeArticle?sfdcid=kA21400d7ZtCAI Our netscaler resource advises that the cookie-based persistence option is not available when he is configuring load-balancing. I believe he is configuring persistence as IP based. Other than that, we can't identify any non-recommended configuration issues. I suspect that the cookie-based load balancer configuration is significant, but not being a LB expert, it's outside of my skillset and my most convincing argument at the moment is, "...because BMC Support recommends it..." Does anyone have any experience with a basic, simple Netscaler configuration to balance two SSL Mid tiers based on least number of connections (or any other strategy)? Or perhaps session timeouts are a known issue for other reasons that I have yet to stumble across? Thanks, -JDHood -- ARSList mailing list ARSList@arslist.org https://mailman.rrr.se/cgi/listinfo/arslist
Re: Load balancing
That status check page actually queries the server to see if any workflow has changed recently. To do that it has to get a working connection from the midtier to a (possibly load balanced) AR Server and from the AR Server to the (possibly load balanced) database and back. You can see all the logs move if you are watching them SQL log(s). One thing… you have to check for the TEXT “200_OK” in the body of the status check web page, not the HTML return code of 200. When the web page comes back it will always be 200 (otherwise you wouldn’t get a web page). The HTML return code of 200 just tells you that the web server is working. Doug > On Jul 12, 2016, at 4:48 PM, LJ LongWing wrote: > > ** > As I understand it, the StatusCheck.jsp makes a few servlet calls to check > for various things. Typically, the server is up, and the Status Check > reports it as up. I have had scenarios where the application server was > entirely off, and was the only app server on the Mid-Tier server, and the > Status Check was reporting as 200_OKI've also experienced where it was > NOT reporting 200_OK when the server was in fact online. I never went into > specifics of why it was reporting the wrong status...I had my own servlet > that I used that gave me everything I needed, accurately, without fail, so I > simply used it instead. I was using Remedy/Mid-Tier 8.1.02 when this > inconsistency was happening. > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
In our flaky 7.6.04 system we would get false failures - the server was up but the OK never happened. That has not happened in our stable 8.1 and 9x versions. It's been really dependable. William Rentfrow wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of LJ LongWing Sent: Tuesday, July 12, 2016 4:48 PM To: arslist@ARSLIST.ORG Subject: Re: Load balancing ** As I understand it, the StatusCheck.jsp makes a few servlet calls to check for various things. Typically, the server is up, and the Status Check reports it as up. I have had scenarios where the application server was entirely off, and was the only app server on the Mid-Tier server, and the Status Check was reporting as 200_OKI've also experienced where it was NOT reporting 200_OK when the server was in fact online. I never went into specifics of why it was reporting the wrong status...I had my own servlet that I used that gave me everything I needed, accurately, without fail, so I simply used it instead. I was using Remedy/Mid-Tier 8.1.02 when this inconsistency was happening. On Tue, Jul 12, 2016 at 3:13 PM, Misi Mladoniczky mailto:m...@rrr.se>> wrote: Hi, Would be interesting to know what the URL actually does. Any API-call to the AR Server? Why did it return the opposite of what it was supposed to show? Best Regards - Misi, RRR AB, http://www.rrr.se (ARSList MVP 2011) Ask the Remedy Licensing Experts (Best R.O.I. Award at WWRUG10/11/12/13): * RRR|License - Not enough Remedy licenses? Save money by optimizing. * RRR|Log - Performance issues or elusive bugs? Analyze your Remedy logs. Find these products, and many free tools and utilities, at http://rrr.se. > I have had personal experience where that URL both showed the service as up > when it wasn't, and down when it wasn'tso, I personally don't trust > that particular url. > > On Tue, Jul 12, 2016 at 1:59 PM, William Rentfrow < > wrentf...@stratacominc.com<mailto:wrentf...@stratacominc.com>> wrote: > >> You can run the back door monitor on the web: >> >> The URL is http://yourUrl/arsys/samples/StatusCheck.jsp >> >> If all is well it returns a very basic page that says: 200_OK >> >> Anything else, you have problems. >> >> William Rentfrow >> wrentf...@stratacominc.com<mailto:wrentf...@stratacominc.com> >> Office: 715-204-3061 or >> 701-232-5697x25 >> Cell: 715-498-5056 >> >> -Original Message- >> From: Action Request System discussion list(ARSList) [mailto: >> arslist@ARSLIST.ORG<mailto:arslist@ARSLIST.ORG>] On Behalf Of Jaffer Mahsoob >> Sent: Monday, July 11, 2016 5:02 PM >> To: arslist@ARSLIST.ORG<mailto:arslist@ARSLIST.ORG> >> Subject: Re: Load balancing >> >> Hey Mark, here's what I've done before... I used the mid-tier sample >> script to check if a specific server is responding to requests, then set >> the netscaller to use the script URL to check the servers, it responds with >> a 200(OK) if the server responds, if not then disable the port redirect >> until it comes back up... It works on Cisco, should be identical >> functionality on Netscaller If you need me to send it to you, let me >> know, you know where to find me... >> >> >> -Jaffer >> >> >> ___ >> UNSUBSCRIBE or access ARSlist Archives at >> www.arslist.org<http://www.arslist.org> >> "Where the Answers Are, and have been for 20 years" >> >> - >> No virus found in this message. >> Checked by AVG - www.avg.com<http://www.avg.com> >> Version: 2016.0.7640 / Virus Database: 4604/12494 - Release Date: 06/25/16 >> Internal Virus Database is out of date. >> >> >> ___ >> UNSUBSCRIBE or access ARSlist Archives at >> www.arslist.org<http://www.arslist.org> >> "Where the Answers Are, and have been for 20 years" >> > > ___ > UNSUBSCRIBE or access ARSlist Archives at > www.arslist.org<http://www.arslist.org> > "Where the Answers Are, and have been for 20 years" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org<http://www.arslist.org> "Where the Answers Are, and have been for 20 years" _ARSlist: "Where the Answers Are" and have been for 20 years_ No virus found in this message. Checked by AVG - www.avg.com<http://www.avg.com> Version: 2016.0.7640 / Virus Database: 4604/12494 - Release Date: 06/25/16 Internal Virus Database is out of date. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
As I understand it, the StatusCheck.jsp makes a few servlet calls to check for various things. Typically, the server is up, and the Status Check reports it as up. I have had scenarios where the application server was entirely off, and was the only app server on the Mid-Tier server, and the Status Check was reporting as 200_OKI've also experienced where it was NOT reporting 200_OK when the server was in fact online. I never went into specifics of why it was reporting the wrong status...I had my own servlet that I used that gave me everything I needed, accurately, without fail, so I simply used it instead. I was using Remedy/Mid-Tier 8.1.02 when this inconsistency was happening. On Tue, Jul 12, 2016 at 3:13 PM, Misi Mladoniczky wrote: > Hi, > > Would be interesting to know what the URL actually does. Any API-call to > the > AR Server? Why did it return the opposite of what it was supposed to show? > > Best Regards - Misi, RRR AB, http://www.rrr.se (ARSList MVP 2011) > > Ask the Remedy Licensing Experts (Best R.O.I. Award at WWRUG10/11/12/13): > * RRR|License - Not enough Remedy licenses? Save money by optimizing. > * RRR|Log - Performance issues or elusive bugs? Analyze your Remedy logs. > Find these products, and many free tools and utilities, at http://rrr.se. > > > I have had personal experience where that URL both showed the service as > up > > when it wasn't, and down when it wasn'tso, I personally don't trust > > that particular url. > > > > On Tue, Jul 12, 2016 at 1:59 PM, William Rentfrow < > > wrentf...@stratacominc.com> wrote: > > > >> You can run the back door monitor on the web: > >> > >> The URL is http://yourUrl/arsys/samples/StatusCheck.jsp > >> > >> If all is well it returns a very basic page that says: 200_OK > >> > >> Anything else, you have problems. > >> > >> William Rentfrow > >> wrentf...@stratacominc.com > >> Office: 715-204-3061 or 701-232-5697x25 > >> Cell: 715-498-5056 > >> > >> -Original Message- > >> From: Action Request System discussion list(ARSList) [mailto: > >> arslist@ARSLIST.ORG] On Behalf Of Jaffer Mahsoob > >> Sent: Monday, July 11, 2016 5:02 PM > >> To: arslist@ARSLIST.ORG > >> Subject: Re: Load balancing > >> > >> Hey Mark, here's what I've done before... I used the mid-tier sample > >> script to check if a specific server is responding to requests, then set > >> the netscaller to use the script URL to check the servers, it responds > with > >> a 200(OK) if the server responds, if not then disable the port redirect > >> until it comes back up... It works on Cisco, should be identical > >> functionality on Netscaller If you need me to send it to you, let > me > >> know, you know where to find me... > >> > >> > >> -Jaffer > >> > >> > >> > ___ > >> UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > >> "Where the Answers Are, and have been for 20 years" > >> > >> - > >> No virus found in this message. > >> Checked by AVG - www.avg.com > >> Version: 2016.0.7640 / Virus Database: 4604/12494 - Release Date: > 06/25/16 > >> Internal Virus Database is out of date. > >> > >> > >> > ___ > >> UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > >> "Where the Answers Are, and have been for 20 years" > >> > > > > > ___ > > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > > "Where the Answers Are, and have been for 20 years" > > > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > "Where the Answers Are, and have been for 20 years" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
I was wondering that too. Maybe it was just making a trip to the web server and back telling the client the web server was up and all that would serve was show the login.jsp page successfully. A trip to the AR Server would be necessary to check for the response of the AR Server as well. This would mean you need to send another signal to the AR Server immediately before or after the status check for the mid-tier using one of the available tools like arsignal or arcreateentry or something simlar.. Joe -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Misi Mladoniczky Sent: Tuesday, July 12, 2016 5:13 PM To: arslist@ARSLIST.ORG Subject: Re: Load balancing Hi, Would be interesting to know what the URL actually does. Any API-call to the AR Server? Why did it return the opposite of what it was supposed to show? Best Regards - Misi, RRR AB, http://www.rrr.se (ARSList MVP 2011) Ask the Remedy Licensing Experts (Best R.O.I. Award at WWRUG10/11/12/13): * RRR|License - Not enough Remedy licenses? Save money by optimizing. * RRR|Log - Performance issues or elusive bugs? Analyze your Remedy logs. Find these products, and many free tools and utilities, at http://rrr.se. > I have had personal experience where that URL both showed the service as up > when it wasn't, and down when it wasn'tso, I personally don't trust > that particular url. > > On Tue, Jul 12, 2016 at 1:59 PM, William Rentfrow < > wrentf...@stratacominc.com> wrote: > >> You can run the back door monitor on the web: >> >> The URL is http://yourUrl/arsys/samples/StatusCheck.jsp >> >> If all is well it returns a very basic page that says: 200_OK >> >> Anything else, you have problems. >> >> William Rentfrow >> wrentf...@stratacominc.com >> Office: 715-204-3061 or 701-232-5697x25 >> Cell: 715-498-5056 >> >> -Original Message- >> From: Action Request System discussion list(ARSList) [mailto: >> arslist@ARSLIST.ORG] On Behalf Of Jaffer Mahsoob >> Sent: Monday, July 11, 2016 5:02 PM >> To: arslist@ARSLIST.ORG >> Subject: Re: Load balancing >> >> Hey Mark, here's what I've done before... I used the mid-tier sample >> script to check if a specific server is responding to requests, then set >> the netscaller to use the script URL to check the servers, it responds with >> a 200(OK) if the server responds, if not then disable the port redirect >> until it comes back up... It works on Cisco, should be identical >> functionality on Netscaller If you need me to send it to you, let me >> know, you know where to find me... >> >> >> -Jaffer >> >> >> ___ >> UNSUBSCRIBE or access ARSlist Archives at www.arslist.org >> "Where the Answers Are, and have been for 20 years" >> >> - >> No virus found in this message. >> Checked by AVG - www.avg.com >> Version: 2016.0.7640 / Virus Database: 4604/12494 - Release Date: 06/25/16 >> Internal Virus Database is out of date. >> >> >> ___ >> UNSUBSCRIBE or access ARSlist Archives at www.arslist.org >> "Where the Answers Are, and have been for 20 years" >> > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > "Where the Answers Are, and have been for 20 years" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years" ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
Hi, Would be interesting to know what the URL actually does. Any API-call to the AR Server? Why did it return the opposite of what it was supposed to show? Best Regards - Misi, RRR AB, http://www.rrr.se (ARSList MVP 2011) Ask the Remedy Licensing Experts (Best R.O.I. Award at WWRUG10/11/12/13): * RRR|License - Not enough Remedy licenses? Save money by optimizing. * RRR|Log - Performance issues or elusive bugs? Analyze your Remedy logs. Find these products, and many free tools and utilities, at http://rrr.se. > I have had personal experience where that URL both showed the service as up > when it wasn't, and down when it wasn'tso, I personally don't trust > that particular url. > > On Tue, Jul 12, 2016 at 1:59 PM, William Rentfrow < > wrentf...@stratacominc.com> wrote: > >> You can run the back door monitor on the web: >> >> The URL is http://yourUrl/arsys/samples/StatusCheck.jsp >> >> If all is well it returns a very basic page that says: 200_OK >> >> Anything else, you have problems. >> >> William Rentfrow >> wrentf...@stratacominc.com >> Office: 715-204-3061 or 701-232-5697x25 >> Cell: 715-498-5056 >> >> -Original Message- >> From: Action Request System discussion list(ARSList) [mailto: >> arslist@ARSLIST.ORG] On Behalf Of Jaffer Mahsoob >> Sent: Monday, July 11, 2016 5:02 PM >> To: arslist@ARSLIST.ORG >> Subject: Re: Load balancing >> >> Hey Mark, here's what I've done before... I used the mid-tier sample >> script to check if a specific server is responding to requests, then set >> the netscaller to use the script URL to check the servers, it responds with >> a 200(OK) if the server responds, if not then disable the port redirect >> until it comes back up... It works on Cisco, should be identical >> functionality on Netscaller If you need me to send it to you, let me >> know, you know where to find me... >> >> >> -Jaffer >> >> >> ___ >> UNSUBSCRIBE or access ARSlist Archives at www.arslist.org >> "Where the Answers Are, and have been for 20 years" >> >> - >> No virus found in this message. >> Checked by AVG - www.avg.com >> Version: 2016.0.7640 / Virus Database: 4604/12494 - Release Date: 06/25/16 >> Internal Virus Database is out of date. >> >> >> ___ >> UNSUBSCRIBE or access ARSlist Archives at www.arslist.org >> "Where the Answers Are, and have been for 20 years" >> > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > "Where the Answers Are, and have been for 20 years" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
I had one older flaky system where it was...well, flaky. But it's pretty consistent for us on ARS 8 and 9. William Rentfrow wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of LJ LongWing Sent: Tuesday, July 12, 2016 3:19 PM To: arslist@ARSLIST.ORG Subject: Re: Load balancing ** I have had personal experience where that URL both showed the service as up when it wasn't, and down when it wasn'tso, I personally don't trust that particular url. On Tue, Jul 12, 2016 at 1:59 PM, William Rentfrow mailto:wrentf...@stratacominc.com>> wrote: You can run the back door monitor on the web: The URL is http://yourUrl/arsys/samples/StatusCheck.jsp If all is well it returns a very basic page that says: 200_OK Anything else, you have problems. William Rentfrow wrentf...@stratacominc.com<mailto:wrentf...@stratacominc.com> Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG<mailto:arslist@ARSLIST.ORG>] On Behalf Of Jaffer Mahsoob Sent: Monday, July 11, 2016 5:02 PM To: arslist@ARSLIST.ORG<mailto:arslist@ARSLIST.ORG> Subject: Re: Load balancing Hey Mark, here's what I've done before... I used the mid-tier sample script to check if a specific server is responding to requests, then set the netscaller to use the script URL to check the servers, it responds with a 200(OK) if the server responds, if not then disable the port redirect until it comes back up... It works on Cisco, should be identical functionality on Netscaller If you need me to send it to you, let me know, you know where to find me... -Jaffer ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org<http://www.arslist.org> "Where the Answers Are, and have been for 20 years" - No virus found in this message. Checked by AVG - www.avg.com<http://www.avg.com> Version: 2016.0.7640 / Virus Database: 4604/12494 - Release Date: 06/25/16 Internal Virus Database is out of date. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org<http://www.arslist.org> "Where the Answers Are, and have been for 20 years" _ARSlist: "Where the Answers Are" and have been for 20 years_ No virus found in this message. Checked by AVG - www.avg.com<http://www.avg.com> Version: 2016.0.7640 / Virus Database: 4604/12494 - Release Date: 06/25/16 Internal Virus Database is out of date. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
I have had personal experience where that URL both showed the service as up when it wasn't, and down when it wasn'tso, I personally don't trust that particular url. On Tue, Jul 12, 2016 at 1:59 PM, William Rentfrow < wrentf...@stratacominc.com> wrote: > You can run the back door monitor on the web: > > The URL is http://yourUrl/arsys/samples/StatusCheck.jsp > > If all is well it returns a very basic page that says: 200_OK > > Anything else, you have problems. > > William Rentfrow > wrentf...@stratacominc.com > Office: 715-204-3061 or 701-232-5697x25 > Cell: 715-498-5056 > > -Original Message- > From: Action Request System discussion list(ARSList) [mailto: > arslist@ARSLIST.ORG] On Behalf Of Jaffer Mahsoob > Sent: Monday, July 11, 2016 5:02 PM > To: arslist@ARSLIST.ORG > Subject: Re: Load balancing > > Hey Mark, here's what I've done before... I used the mid-tier sample > script to check if a specific server is responding to requests, then set > the netscaller to use the script URL to check the servers, it responds with > a 200(OK) if the server responds, if not then disable the port redirect > until it comes back up... It works on Cisco, should be identical > functionality on Netscaller If you need me to send it to you, let me > know, you know where to find me... > > > -Jaffer > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > "Where the Answers Are, and have been for 20 years" > > - > No virus found in this message. > Checked by AVG - www.avg.com > Version: 2016.0.7640 / Virus Database: 4604/12494 - Release Date: 06/25/16 > Internal Virus Database is out of date. > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > "Where the Answers Are, and have been for 20 years" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
You can run the back door monitor on the web: The URL is http://yourUrl/arsys/samples/StatusCheck.jsp If all is well it returns a very basic page that says: 200_OK Anything else, you have problems. William Rentfrow wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Jaffer Mahsoob Sent: Monday, July 11, 2016 5:02 PM To: arslist@ARSLIST.ORG Subject: Re: Load balancing Hey Mark, here's what I've done before... I used the mid-tier sample script to check if a specific server is responding to requests, then set the netscaller to use the script URL to check the servers, it responds with a 200(OK) if the server responds, if not then disable the port redirect until it comes back up... It works on Cisco, should be identical functionality on Netscaller If you need me to send it to you, let me know, you know where to find me... -Jaffer ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years" - No virus found in this message. Checked by AVG - www.avg.com Version: 2016.0.7640 / Virus Database: 4604/12494 - Release Date: 06/25/16 Internal Virus Database is out of date. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
Hey Mark, here's what I've done before... I used the mid-tier sample script to check if a specific server is responding to requests, then set the netscaller to use the script URL to check the servers, it responds with a 200(OK) if the server responds, if not then disable the port redirect until it comes back up... It works on Cisco, should be identical functionality on Netscaller If you need me to send it to you, let me know, you know where to find me... -Jaffer ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
I was wondering if you could copy the arsignal and its supporting libraries where needed and use them to get a signal back on whether or not the AR System server is up or not responding.. Or use a utility like arcreateentry.. Joe _ From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of laurent matheo Sent: Monday, July 11, 2016 10:42 AM To: arslist@ARSLIST.ORG Subject: Re: Load balancing Hi Rick :) What is the life test implemented on the Netscaler? If it's "just" the ARS TCP port ping it will not be enough as LJ mentioned and you will have the same problems using the BigIP F5. I was recently at a customer site where the F5 did not detect that the ARS was out due to the "ping" on the ARS TCP port. Indeed there are cases where an ARS server can just be "out" (not handling requests) but continue to accept connection on the TCP port. As LJ said, one of the option is to write a "real" life test using ARS Java APIs (for example): -> Login, -> Query on COM:Company to search for "- Global -", -> Logout, -> It would write a file somewhere, -> This file would be parsed by the Load Balancer to check if the ARS Server is answering requests and to exclude or include it from the Load Balancer, Best regards. Laurent. On 11 Jul, 2016, at 04:12 PM, Rick Cook wrote: ** Mark, FWIW, we are seeing issues with Netscalers as well, on an 8.1 platform. Servers that are part and running get dropped from the LB, etc. We've been working with our system engineers on it, but it may be a limitation in the Netscaler architecture. We asked for F5s Rick On Jul 11, 2016 06:49, "Brittain, Mark" wrote: ** Hi All I have two mid-tiers and two AR Servers in a server group with netscaler load balancing in-between. Had a situation where the second AR Server in the group stopped responding but the load balancing did not detect this and whole server group became inaccessible. Digging a little deeper into the load balancing determined the netscaler pings the per port but doesn't go any deeper. So in this situation, the port was up and netscaler continued to push traffic to the hung server. Is there something better that netscaler can monitor that would detect when the AR Server stops? If not, is there any way to ensure that if one server quits, everything goes to the other server and users are not interrupted? ARS 7.6.04 SP5 ITSM 7.6.04 SP2 Mid-Tier 7.6.04 SP5 Thanks Mark Mark Brittain | Sr. Systems Engineer | 315.637.9337 O 315.882.5360 M 125 Elwood Davis Road | Syracuse NY 13212 Navisite _ This E-mail and any of its attachments may contain Time Warner Cable proprietary information, which is privileged, confidential, or subject to copyright belonging to Time Warner Cable. This E-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this E-mail, you are hereby notified that any dissemination, distribution, copying, or action taken in relation to the contents of and attachments to this E-mail is strictly prohibited and may be unlawful. If you have received this E-mail in error, please notify the sender immediately and permanently delete the original and any copy of this E-mail and any printout. _ARSlist: "Where the Answers Are" and have been for 20 years_ _ARSlist: "Where the Answers Are" and have been for 20 years_ _ARSlist: "Where the Answers Are" and have been for 20 years_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
Hi Rick :) What is the life test implemented on the Netscaler? If it's "just" the ARS TCP port ping it will not be enough as LJ mentioned and you will have the same problems using the BigIP F5. I was recently at a customer site where the F5 did not detect that the ARS was out due to the "ping" on the ARS TCP port. Indeed there are cases where an ARS server can just be "out" (not handling requests) but continue to accept connection on the TCP port. As LJ said, one of the option is to write a "real" life test using ARS Java APIs (for example): -> Login, -> Query on COM:Company to search for "- Global -", -> Logout, -> It would write a file somewhere, -> This file would be parsed by the Load Balancer to check if the ARS Server is answering requests and to exclude or include it from the Load Balancer, Best regards. Laurent. On 11 Jul, 2016, at 04:12 PM, Rick Cook wrote: ** Mark, FWIW, we are seeing issues with Netscalers as well, on an 8.1 platform. Servers that are part and running get dropped from the LB, etc. We've been working with our system engineers on it, but it may be a limitation in the Netscaler architecture. We asked for F5s Rick On Jul 11, 2016 06:49, "Brittain, Mark" wrote: ** Hi All I have two mid-tiers and two AR Servers in a server group with netscaler load balancing in-between. Had a situation where the second AR Server in the group stopped responding but the load balancing did not detect this and whole server group became inaccessible. Digging a little deeper into the load balancing determined the netscaler pings the per port but doesn’t go any deeper. So in this situation, the port was up and netscaler continued to push traffic to the hung server. Is there something better that netscaler can monitor that would detect when the AR Server stops? If not, is there any way to ensure that if one server quits, everything goes to the other server and users are not interrupted? ARS 7.6.04 SP5 ITSM 7.6.04 SP2 Mid-Tier 7.6.04 SP5 Thanks Mark Mark Brittain | Sr. Systems Engineer | 315.637.9337 O 315.882.5360 M 125 Elwood Davis Road | Syracuse NY 13212 This E-mail and any of its attachments may contain Time Warner Cable proprietary information, which is privileged, confidential, or subject to copyright belonging to Time Warner Cable. This E-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this E-mail, you are hereby notified that any dissemination, distribution, copying, or action taken in relation to the contents of and attachments to this E-mail is strictly prohibited and may be unlawful. If you have received this E-mail in error, please notify the sender immediately and permanently delete the original and any copy of this E-mail and any printout. _ARSlist: "Where the Answers Are" and have been for 20 years_ _ARSlist: "Where the Answers Are" and have been for 20 years_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
Mark, FWIW, we are seeing issues with Netscalers as well, on an 8.1 platform. Servers that are part and running get dropped from the LB, etc. We've been working with our system engineers on it, but it may be a limitation in the Netscaler architecture. We asked for F5s Rick On Jul 11, 2016 06:49, "Brittain, Mark" wrote: ** Hi All I have two mid-tiers and two AR Servers in a server group with netscaler load balancing in-between. Had a situation where the second AR Server in the group stopped responding but the load balancing did not detect this and whole server group became inaccessible. Digging a little deeper into the load balancing determined the netscaler pings the per port but doesn’t go any deeper. So in this situation, the port was up and netscaler continued to push traffic to the hung server. Is there something better that netscaler can monitor that would detect when the AR Server stops? If not, is there any way to ensure that if one server quits, everything goes to the other server and users are not interrupted? ARS 7.6.04 SP5 ITSM 7.6.04 SP2 Mid-Tier 7.6.04 SP5 Thanks Mark *Mark Brittain * | Sr. Systems Engineer | 315.637.9337 O 315.882.5360 M 125 Elwood Davis Road | Syracuse NY 13212 [image: Navisite] -- This E-mail and any of its attachments may contain Time Warner Cable proprietary information, which is privileged, confidential, or subject to copyright belonging to Time Warner Cable. This E-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this E-mail, you are hereby notified that any dissemination, distribution, copying, or action taken in relation to the contents of and attachments to this E-mail is strictly prohibited and may be unlawful. If you have received this E-mail in error, please notify the sender immediately and permanently delete the original and any copy of this E-mail and any printout. _ARSlist: "Where the Answers Are" and have been for 20 years_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load balancing
Mark, Smart Monitoring from the LB is always a critical thing, yet the discussion doesn't usually start till you have your exact situation and people start looking into 'what can the LB do to determine if a web/app node is ACTUALLY online'. I've dealt with this in the past. My solution for it was to write a program that 'exercised' the Remedy system, log on, do some queries, etc...whatever it is you feel is appropriate to verify that the node is working properly. Have that program 'do something' that the LB can use to determine 'up or down'as you found, port monitoring doesn't work, not on the Remedy process at leastdo you have any quality programmers that can exercise Remedy server via API, and do something like put a file on the filesystem that the LB can look for? On Mon, Jul 11, 2016 at 7:12 AM, Brittain, Mark wrote: > ** > > Hi All > > > > I have two mid-tiers and two AR Servers in a server group with netscaler > load balancing in-between. Had a situation where the second AR Server in > the group stopped responding but the load balancing did not detect this and > whole server group became inaccessible. Digging a little deeper into the > load balancing determined the netscaler pings the per port but doesn’t go > any deeper. So in this situation, the port was up and netscaler continued > to push traffic to the hung server. > > > > Is there something better that netscaler can monitor that would detect > when the AR Server stops? If not, is there any way to ensure that if one > server quits, everything goes to the other server and users are not > interrupted? > > > > ARS 7.6.04 SP5 > > ITSM 7.6.04 SP2 > > Mid-Tier 7.6.04 SP5 > > > > Thanks > > Mark > > > > *Mark Brittain * | Sr. Systems Engineer | 315.637.9337 O 315.882.5360 M > > 125 Elwood Davis Road | Syracuse NY 13212 > > [image: Navisite] > > > > -- > > This E-mail and any of its attachments may contain Time Warner Cable > proprietary information, which is privileged, confidential, or subject to > copyright belonging to Time Warner Cable. This E-mail is intended solely > for the use of the individual or entity to which it is addressed. If you > are not the intended recipient of this E-mail, you are hereby notified that > any dissemination, distribution, copying, or action taken in relation to > the contents of and attachments to this E-mail is strictly prohibited and > may be unlawful. If you have received this E-mail in error, please notify > the sender immediately and permanently delete the original and any copy of > this E-mail and any printout. > _ARSlist: "Where the Answers Are" and have been for 20 years_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Load balancing
Hi All I have two mid-tiers and two AR Servers in a server group with netscaler load balancing in-between. Had a situation where the second AR Server in the group stopped responding but the load balancing did not detect this and whole server group became inaccessible. Digging a little deeper into the load balancing determined the netscaler pings the per port but doesn't go any deeper. So in this situation, the port was up and netscaler continued to push traffic to the hung server. Is there something better that netscaler can monitor that would detect when the AR Server stops? If not, is there any way to ensure that if one server quits, everything goes to the other server and users are not interrupted? ARS 7.6.04 SP5 ITSM 7.6.04 SP2 Mid-Tier 7.6.04 SP5 Thanks Mark Mark Brittain | Sr. Systems Engineer | 315.637.9337 O 315.882.5360 M 125 Elwood Davis Road | Syracuse NY 13212 [Navisite] This E-mail and any of its attachments may contain Time Warner Cable proprietary information, which is privileged, confidential, or subject to copyright belonging to Time Warner Cable. This E-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this E-mail, you are hereby notified that any dissemination, distribution, copying, or action taken in relation to the contents of and attachments to this E-mail is strictly prohibited and may be unlawful. If you have received this E-mail in error, please notify the sender immediately and permanently delete the original and any copy of this E-mail and any printout. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Load Balancing on Remedy Server Group
Raj, For your environment, things are pretty simple. You have two servers you want to load balance across. You can use any load balancer mechanism you want to simply direct traffic to either of them using whatever balancing strategy (round robin, percentage, whatever) you want. You can use a software load balancer without issue – in fact, the majority of AR System customers do use software load balancers. It can run on the same machine as one of the AR System servers or on a separate machine (of course if you run on one of the same machines, if you lose the machine, you lose your load balancer so it is best to run on a separate machine). And, wherever I say machine, it can be a VM of course. There is no special configuration you have to worry about for a simple scenario of a balancer over AR System servers. Doug Mueller From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Raj Sent: Monday, December 29, 2014 7:15 PM To: arslist@ARSLIST.ORG Subject: Load Balancing on Remedy Server Group ** Hello All, Looking for some information on load balancing our remedy in server group install. Currently, we installed ARS v8.1 in Server Group on 2 VMs(OS - Linux). We tested some fail-over scenarios and it is working fine after we configured entries in Server Group Operation Ranking form. We are fairly new to Server Group install and load balancing our Remedy environment. We do not have mid-tier but we only use Remedy User Tool(v7.6.04 SP5). We may not be able to go the route of hardware load balancing since it would require new hardware/server provision. Considering the above facts, looking for some suggestions/recommendations from Remedy gurus in here on what can be done in our environment to facilitate load balancing. Regards, Raj _ARSlist: "Where the Answers Are" and have been for 20 years_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Load Balancing on Remedy Server Group
Hello All, Looking for some information on load balancing our remedy in server group install. Currently, we installed ARS v8.1 in Server Group on 2 VMs(OS - Linux). We tested some fail-over scenarios and it is working fine after we configured entries in Server Group Operation Ranking form. We are fairly new to Server Group install and load balancing our Remedy environment. We do not have mid-tier but we only use Remedy User Tool(v7.6.04 SP5). We may not be able to go the route of hardware load balancing since it would require new hardware/server provision. Considering the above facts, looking for some suggestions/recommendations from Remedy gurus in here on what can be done in our environment to facilitate load balancing. Regards, Raj ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Load balancing ARS 81 using F5 (posting on behalf of Sabyson Fernandes)
I'm posting this on behalf of my buddy Sabyson who is having a small issue with configuring the F5 Load balancer between the AR Server to the end users browsers.. If any of you have dealt with a similar problem, your response would be appreciated. While I can forward your responses to Sabyson, you could write directly to him at sebyr...@yahoo.com. Seems he has done everything right.. So I can't think of why they should have this problem either. Joe Sabyson's Email= Joe, Help me to post to the list. I got unsubscribed and no time to subscribe. Having issue with Remedy 8.1 IE ==> F5 LB (Sticky enabled) ==> 4 MT Servers (Lifespan enabled) ==> F5 LB (Sticky disabled) ==> 4 ARS (Server group) User connects fine initially. After sometime user gets the error "Connected from another machine". We have the session timeout on Mid-Tier configured at 60 minutes and the TCP Idle session on the F5 set at 62 minutes. ARS 81, MT 8.1, ITSM 8.1 Have a ticket open with BMC but no luck so far. Maybe someone on the list has already resolved this or can provide their F5 settings if they have a similar setup. Thanks, Saby ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Server Group Environment without Load Balancing
I would suggest re-reading the documentation on high availability. Without a load balancer, your users will be directed to a specific node. If that node fails, the users will experience an outage. Remedy does not act as a proxy to distribute the load across servers for end users requests. The server group ranking form let's you designate certain systems for certain server side functions (e.g., escalations, email, approval); this does not apply to end users. If you want to balance your load on either the AR tier or the web tier, you will need to use a load balancer. I suppose you could use DNS round robin for the ARServer tier to distribute the load, but this will do nothing in terms of fault tolerance. Typical load balancer acts as a reverse proxy (a front-man for incoming requests). It verifies the availability of the back-end systems and redirects traffic based on current load distribution and availability. Axton On Wed, Mar 12, 2014 at 9:29 AM, Sivarama Velicheti wrote: > Hi Group, > > We have a server group environment but not leveraging load > balancing capabilities. Here is a brief overview on the production servers > we have in place. I am trying to understand or find some documentation to > best utilize the servers we have at hand. We have 3 application servers in > which remedy is installed (server group environment) i) AppServerA ii) > AppServerB and iii) AppServerC . We further have two webservers for > midtier i) webServerA and ii) webServerB. Here are somethings I am trying > to wrap my head around. If we want to set Domain Name Alias am I correct in > assuming we need to only set it for the primary server AppServerA (since no > load balancer) and use the Operations Group ranking form to manage our > load? We have two webservers and we need to set up the DNS alias for them > as well. I have no idea as to how to use both of them in our environment > without a load balancer. Any ideas or links to any documentation would be > greatly appreciated. I went through the installing and configuring Server > groups guide but have no ideas. We are on Remedy 8.1, Windows environment. > > Thanks > Siva > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > "Where the Answers Are, and have been for 20 years" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Server Group Environment without Load Balancing
Siva, You have three real options. 1 - Use a smart load balancer that can send traffic smartly to an available node with algorithms to manage things properly. 2 - Use DNS 'Round Robin' type of configurationbut this would only work in the App, because the Web needs what's called 'session affinity, or sticky bits' to work properly 3 - Have separate 'names' for the separate servers, and access them individually. >From a load balancer perspective, you can use something as simple as an Apache server to do the load balancing...it's not as quick as hardware load balancing, but it IS load balancing...I don't have any experience with setup and configuration of one, but I imagine it's not exceedingly difficult :) On Wed, Mar 12, 2014 at 8:29 AM, Sivarama Velicheti wrote: > Hi Group, > > We have a server group environment but not leveraging load > balancing capabilities. Here is a brief overview on the production servers > we have in place. I am trying to understand or find some documentation to > best utilize the servers we have at hand. We have 3 application servers in > which remedy is installed (server group environment) i) AppServerA ii) > AppServerB and iii) AppServerC . We further have two webservers for > midtier i) webServerA and ii) webServerB. Here are somethings I am trying > to wrap my head around. If we want to set Domain Name Alias am I correct in > assuming we need to only set it for the primary server AppServerA (since no > load balancer) and use the Operations Group ranking form to manage our > load? We have two webservers and we need to set up the DNS alias for them > as well. I have no idea as to how to use both of them in our environment > without a load balancer. Any ideas or links to any documentation would be > greatly appreciated. I went through the installing and configuring Server > groups guide but have no ideas. We are on Remedy 8.1, Windows environment. > > Thanks > Siva > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > "Where the Answers Are, and have been for 20 years" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Server Group Environment without Load Balancing
Hi Group, We have a server group environment but not leveraging load balancing capabilities. Here is a brief overview on the production servers we have in place. I am trying to understand or find some documentation to best utilize the servers we have at hand. We have 3 application servers in which remedy is installed (server group environment) i) AppServerA ii) AppServerB and iii) AppServerC . We further have two webservers for midtier i) webServerA and ii) webServerB. Here are somethings I am trying to wrap my head around. If we want to set Domain Name Alias am I correct in assuming we need to only set it for the primary server AppServerA (since no load balancer) and use the Operations Group ranking form to manage our load? We have two webservers and we need to set up the DNS alias for them as well. I have no idea as to how to use both of them in our environment without a load balancer. Any ideas or links to any documentation would be greatly appreciated. I went through the installing and configuring Server groups guide but have no ideas. We are on Remedy 8.1, Windows environment. Thanks Siva ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Remedy Load Balancing (UNCLASSIFIED)
Classification: UNCLASSIFIED Caveats: NONE We are running 4 Mid-Tiers behind a hardware load balancer. The mid-tiers are running tomcat 7 with SSL enabled and CAC authentication through Atrium SSO. When users hit the load balancer URL the browser address bar changes to the URL of the mid-tier they get pointed to. I've been told the browser address bar should not change regardless of which mid-tier the user is pointed to. Seems to be some sort of redirection going on within Tomcat on the mid-tier. Is there a configuration somewhere to stop this? If so, can someone tell me what to change? Thanks Rod Griffin Lockheed Martin Remedy Development and Support Classification: UNCLASSIFIED Caveats: NONE ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Remedy Load Balancing (UNCLASSIFIED)
That is something in the VIP settings on your load balancer. We have a friendly DNS name for our mid-tiers that points to the VIP and that is what remains displayed in the address bar. I wrote the server name in the individual login.jsp files so I can View Source on the login screen to verify which server I am hitting. -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Griffin, Rod E ACE-IT Contractor Sent: Friday, September 20, 2013 11:24 AM To: arslist@ARSLIST.ORG Subject: Remedy Load Balancing (UNCLASSIFIED) Classification: UNCLASSIFIED Caveats: NONE We are running 4 Mid-Tiers behind a hardware load balancer. The mid-tiers are running tomcat 7 with SSL enabled and CAC authentication through Atrium SSO. When users hit the load balancer URL the browser address bar changes to the URL of the mid-tier they get pointed to. I've been told the browser address bar should not change regardless of which mid-tier the user is pointed to. Seems to be some sort of redirection going on within Tomcat on the mid-tier. Is there a configuration somewhere to stop this? If so, can someone tell me what to change? Thanks Rod Griffin Lockheed Martin Remedy Development and Support Classification: UNCLASSIFIED Caveats: NONE ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years" ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: Remedy Load Balancing (UNCLASSIFIED)
Classification: UNCLASSIFIED Caveats: NONE Thanks Tommy, I've reached out to our load balancer guy. -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Tommy Morris Sent: Friday, September 20, 2013 11:32 AM To: arslist@ARSLIST.ORG Subject: [EXTERNAL] Re: Remedy Load Balancing (UNCLASSIFIED) That is something in the VIP settings on your load balancer. We have a friendly DNS name for our mid-tiers that points to the VIP and that is what remains displayed in the address bar. I wrote the server name in the individual login.jsp files so I can View Source on the login screen to verify which server I am hitting. -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Griffin, Rod E ACE-IT Contractor Sent: Friday, September 20, 2013 11:24 AM To: arslist@ARSLIST.ORG Subject: Remedy Load Balancing (UNCLASSIFIED) Classification: UNCLASSIFIED Caveats: NONE We are running 4 Mid-Tiers behind a hardware load balancer. The mid-tiers are running tomcat 7 with SSL enabled and CAC authentication through Atrium SSO. When users hit the load balancer URL the browser address bar changes to the URL of the mid-tier they get pointed to. I've been told the browser address bar should not change regardless of which mid-tier the user is pointed to. Seems to be some sort of redirection going on within Tomcat on the mid-tier. Is there a configuration somewhere to stop this? If so, can someone tell me what to change? Thanks Rod Griffin Lockheed Martin Remedy Development and Support Classification: UNCLASSIFIED Caveats: NONE ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years" ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years" Classification: UNCLASSIFIED Caveats: NONE ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
Re: MS Network Load Balancing (NLB) with AR Server Groups
Karthik, Very elegant. I guess it sums up my experiences in general. NLB will work just fine if you don't have issues with ANY node and you simply want to balance between existing nodes. As long as something goes wrong on a node, you have no ability to automatically pull it out of the balance unless the service is stopped, or the machine is offline. -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Karthik Sent: Monday, September 10, 2012 11:45 AM To: arslist@ARSLIST.ORG Subject: Re: MS Network Load Balancing (NLB) with AR Server Groups ** Hi, I just implemented NLB for server groups. NLB was implemented for Mid-tiers and also for App Server group. As someone has mentioned before, its does not provide all the different modes that other hardware load balancers do. only option is for load balancing based on the server load. However, there are some problems with NLB. For Ex: If Mid Tier 1 is down but the server hosting it still provides a heartbeat, clients are still sent to it. This poses a problem when it comes to balancing load to mid tier servers. But for app servers this can be handled using OOB provided configuration, "Enable lifespan" on mid tier config page. By enabling this, mid-tiers will understand if an app server is down and the requests will not be sent to the app servers. however, if the app servers have performance problems and are services are still up, requests will still be sent to them. Only way to mitigate the negative scenarios described above would be to stop a particular host from the NLB manager, while you are debugging the problem with the mid-tier being down/app server having performance issues. that said, below are some inportant configs you need to keep in mind: 1- load balancer should be configured in Multi cast mode if its NLB. Network guys should understand this. 2 - affinity should be set to None. 3 - load balancer name should be FQDN I.e. fully qualified domain name. Regarding point 2 above: this is applicable for only application load balancer. For mid tier load balancer(if you are planning to have one in place) the affinity parameter should be set to single. Sometimes, NLB is configured in unicast mode. in this case, there is a need to add an additional network adapter and this causes license issues as in unicast mode, the NLB makes the mac address of all the NLB hosts the same. to mitigate this, please follow the attached link to not allow the mac address to be the same for all the hosts: http://technet.microsoft.com/en-us/library/bb742455.aspx <http://technet.microsoft.com/en-us/library/bb742455.aspx> hope this helps. My personal and also per the article found in above link, its good to go for multi-cast mode. Regards, Karthik On 10 September 2012 22:49, Mauricio M. wrote: ** Hello, thank you, what would be some of those specific configurations or capabilities? -Mauricio 2012/9/10 Longwing, LJ CTR MDA/IC Mauricio, I recently re-looked at MS NLB and found it to be 'lacking' of proper load balancing configuration capabilities, but I haven't looked real close at it either. I much prefer an independent LB tool. -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mauricio M. Sent: Monday, September 10, 2012 11:06 AM To: arslist@ARSLIST.ORG Subject: MS Network Load Balancing (NLB) with AR Server Groups ** Hello, Does anyone have any experience or feedback using Microsoft NLB for load balancing with AR System 7.6.04? There are a few old posts about NLB but nothing recent Thank you in advance, Mauricio _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are" _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are"
Re: MS Network Load Balancing (NLB) with AR Server Groups
Completely agree with you! Regards, Karthik On 10 September 2012 23:27, Longwing, LJ CTR MDA/IC wrote: > Well...one thing that I have always looked for in a load balancing is a > 'smart' capability to tell you that something is functional. I have found > most of the balancers on the market to not allow what I consider smart > balancing. Most will do a ping, open a port, check a URL, that sort of > thing, but in a Remedy world just because ARServer allows you to open the > port it is on, doesn't mean that the Remedy is functional. Just because > the Mid-Tier login page opens doesn't mean that it's functional. I have > found that some LB suites have various scripting capabilities, but in > general I have found creating API based monitoring tools that provide 'port > opening' capabilities to be a 'full' solution. The general capabilities > that I'm referring to are session affinity (NLB has this apparently). The > rest are probes, the ability to probe your intended subject to see if the > capabilities you are looking for are functional. As previously discussed, > most of these probes are 'dumb', but they are smarter than nothing. NLB > doesn't seem to have any probe capability, the simple 'am I on' is enough > to route traffic to the node. I personally think that it's important to > probe deeply into the application to know that it's not only on, but > functional before routing traffic, and NLB doesn't meet this probing > requirement. > > -Original Message- > From: Action Request System discussion list(ARSList) [mailto: > arslist@ARSLIST.ORG] On Behalf Of Mauricio M. > Sent: Monday, September 10, 2012 11:20 AM > To: arslist@ARSLIST.ORG > Subject: Re: MS Network Load Balancing (NLB) with AR Server Groups > > ** Hello, > > thank you, what would be some of those specific configurations or > capabilities? > > -Mauricio > > > 2012/9/10 Longwing, LJ CTR MDA/IC > > > Mauricio, > I recently re-looked at MS NLB and found it to be 'lacking' of > proper load balancing configuration capabilities, but I haven't looked real > close at it either. I much prefer an independent LB tool. > > -Original Message- > From: Action Request System discussion list(ARSList) [mailto: > arslist@ARSLIST.ORG] On Behalf Of Mauricio M. > Sent: Monday, September 10, 2012 11:06 AM > To: arslist@ARSLIST.ORG > Subject: MS Network Load Balancing (NLB) with AR Server Groups > > ** Hello, > > > Does anyone have any experience or feedback using Microsoft NLB > for load balancing with AR System 7.6.04? > > There are a few old posts about NLB but nothing recent > > Thank you in advance, > > Mauricio > > _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are" > > > > _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are"
Re: MS Network Load Balancing (NLB) with AR Server Groups
Well...one thing that I have always looked for in a load balancing is a 'smart' capability to tell you that something is functional. I have found most of the balancers on the market to not allow what I consider smart balancing. Most will do a ping, open a port, check a URL, that sort of thing, but in a Remedy world just because ARServer allows you to open the port it is on, doesn't mean that the Remedy is functional. Just because the Mid-Tier login page opens doesn't mean that it's functional. I have found that some LB suites have various scripting capabilities, but in general I have found creating API based monitoring tools that provide 'port opening' capabilities to be a 'full' solution. The general capabilities that I'm referring to are session affinity (NLB has this apparently). The rest are probes, the ability to probe your intended subject to see if the capabilities you are looking for are functional. As previously discussed, most of these probes are 'dumb', but they are smarter than nothing. NLB doesn't seem to have any probe capability, the simple 'am I on' is enough to route traffic to the node. I personally think that it's important to probe deeply into the application to know that it's not only on, but functional before routing traffic, and NLB doesn't meet this probing requirement. -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mauricio M. Sent: Monday, September 10, 2012 11:20 AM To: arslist@ARSLIST.ORG Subject: Re: MS Network Load Balancing (NLB) with AR Server Groups ** Hello, thank you, what would be some of those specific configurations or capabilities? -Mauricio 2012/9/10 Longwing, LJ CTR MDA/IC Mauricio, I recently re-looked at MS NLB and found it to be 'lacking' of proper load balancing configuration capabilities, but I haven't looked real close at it either. I much prefer an independent LB tool. -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mauricio M. Sent: Monday, September 10, 2012 11:06 AM To: arslist@ARSLIST.ORG Subject: MS Network Load Balancing (NLB) with AR Server Groups ** Hello, Does anyone have any experience or feedback using Microsoft NLB for load balancing with AR System 7.6.04? There are a few old posts about NLB but nothing recent Thank you in advance, Mauricio _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are" _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are"
Re: MS Network Load Balancing (NLB) with AR Server Groups
Hi, I just implemented NLB for server groups. NLB was implemented for Mid-tiers and also for App Server group. As someone has mentioned before, its does not provide all the different modes that other hardware load balancers do. only option is for load balancing based on the server load. However, there are some problems with NLB. For Ex: If Mid Tier 1 is down but the server hosting it still provides a heartbeat, clients are still sent to it. This poses a problem when it comes to balancing load to mid tier servers. But for app servers this can be handled using OOB provided configuration, "Enable lifespan" on mid tier config page. By enabling this, mid-tiers will understand if an app server is down and the requests will not be sent to the app servers. however, if the app servers have performance problems and are services are still up, requests will still be sent to them. Only way to mitigate the negative scenarios described above would be to stop a particular host from the NLB manager, while you are debugging the problem with the mid-tier being down/app server having performance issues. that said, below are some inportant configs you need to keep in mind: 1- load balancer should be configured in Multi cast mode if its NLB. Network guys should understand this. 2 - affinity should be set to None. 3 - load balancer name should be FQDN I.e. fully qualified domain name. Regarding point 2 above: this is applicable for only application load balancer. For mid tier load balancer(if you are planning to have one in place) the affinity parameter should be set to single. Sometimes, NLB is configured in unicast mode. in this case, there is a need to add an additional network adapter and this causes license issues as in unicast mode, the NLB makes the mac address of all the NLB hosts the same. to mitigate this, please follow the attached link to not allow the mac address to be the same for all the hosts: *http://technet.microsoft.com/en-us/library/bb742455.aspx * hope this helps. My personal and also per the article found in above link, its good to go for multi-cast mode. Regards, Karthik On 10 September 2012 22:49, Mauricio M. wrote: > ** Hello, > > thank you, what would be some of those specific configurations or > capabilities? > > -Mauricio > > 2012/9/10 Longwing, LJ CTR MDA/IC > > Mauricio, >> I recently re-looked at MS NLB and found it to be 'lacking' of proper >> load balancing configuration capabilities, but I haven't looked real close >> at it either. I much prefer an independent LB tool. >> >> -Original Message- >> From: Action Request System discussion list(ARSList) [mailto: >> arslist@ARSLIST.ORG] On Behalf Of Mauricio M. >> Sent: Monday, September 10, 2012 11:06 AM >> To: arslist@ARSLIST.ORG >> Subject: MS Network Load Balancing (NLB) with AR Server Groups >> >> ** Hello, >> >> Does anyone have any experience or feedback using Microsoft NLB for load >> balancing with AR System 7.6.04? >> >> There are a few old posts about NLB but nothing recent >> >> Thank you in advance, >> >> Mauricio >> _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ >> >> >> ___ >> UNSUBSCRIBE or access ARSlist Archives at www.arslist.org >> attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are" >> > > _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are"
Re: MS Network Load Balancing (NLB) with AR Server Groups
Hello, thank you, what would be some of those specific configurations or capabilities? -Mauricio 2012/9/10 Longwing, LJ CTR MDA/IC > Mauricio, > I recently re-looked at MS NLB and found it to be 'lacking' of proper load > balancing configuration capabilities, but I haven't looked real close at it > either. I much prefer an independent LB tool. > > -Original Message- > From: Action Request System discussion list(ARSList) [mailto: > arslist@ARSLIST.ORG] On Behalf Of Mauricio M. > Sent: Monday, September 10, 2012 11:06 AM > To: arslist@ARSLIST.ORG > Subject: MS Network Load Balancing (NLB) with AR Server Groups > > ** Hello, > > Does anyone have any experience or feedback using Microsoft NLB for load > balancing with AR System 7.6.04? > > There are a few old posts about NLB but nothing recent > > Thank you in advance, > > Mauricio > _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are"
Re: MS Network Load Balancing (NLB) with AR Server Groups
Mauricio, I recently re-looked at MS NLB and found it to be 'lacking' of proper load balancing configuration capabilities, but I haven't looked real close at it either. I much prefer an independent LB tool. -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mauricio M. Sent: Monday, September 10, 2012 11:06 AM To: arslist@ARSLIST.ORG Subject: MS Network Load Balancing (NLB) with AR Server Groups ** Hello, Does anyone have any experience or feedback using Microsoft NLB for load balancing with AR System 7.6.04? There are a few old posts about NLB but nothing recent Thank you in advance, Mauricio _attend WWRUG12 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are"
MS Network Load Balancing (NLB) with AR Server Groups
Hello, Does anyone have any experience or feedback using Microsoft NLB for load balancing with AR System 7.6.04? There are a few old posts about NLB but nothing recent Thank you in advance, Mauricio ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug12 www.wwrug12.com ARSList: "Where the Answers Are"
Re: Apache Front End configuration for web logic load balancing
There is also mod_proxy_ajp in Apache 2.2 that allows you to use AJP without the problems in mod_jk. mod_proxy_ajp can be compiled for Apache 2.0 for those folks running that version of Apache. AJP, in practice, is more efficient than using HTTP. Axton On Wed, Sep 21, 2011 at 10:22 PM, John Baker wrote: > Hello, > > Chuck: You are referring to session replication. I do not believe BMC > support this although it may work. If all the objects placed in the session > are serializable then it should work. > > Sean: mod_wl will perform the load balancing for you, but WL has an HTTP > connector so if Apache is doing nothing else but serving Mid Tier then it's > worth considering dropping them completely in order to simplify the > configuration. > > For completeness, if I prenteded you were running Tomct, the LB between > Apache and Mid Tier is not what would be considered a standard deployment > for Apache/Tomcat. If you use mod_jk (but don't), it will handle the LB > between MT instances for you. If you use mod_proxy or mod_cluster, it'll > also handle the LB (better options than mod_jk). > > The only piece I can spot missing from your configuration is SSO Plugin. > :-) > > > John > > -- > John Baker, Web Technologies Consultant / SSO Support. > http://www.javasystemsolutions.com/jss/ssoplugin > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Apache Front End configuration for web logic load balancing
Hello, Chuck: You are referring to session replication. I do not believe BMC support this although it may work. If all the objects placed in the session are serializable then it should work. Sean: mod_wl will perform the load balancing for you, but WL has an HTTP connector so if Apache is doing nothing else but serving Mid Tier then it's worth considering dropping them completely in order to simplify the configuration. For completeness, if I prenteded you were running Tomct, the LB between Apache and Mid Tier is not what would be considered a standard deployment for Apache/Tomcat. If you use mod_jk (but don't), it will handle the LB between MT instances for you. If you use mod_proxy or mod_cluster, it'll also handle the LB (better options than mod_jk). The only piece I can spot missing from your configuration is SSO Plugin. :-) John -- John Baker, Web Technologies Consultant / SSO Support. http://www.javasystemsolutions.com/jss/ssoplugin ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: Apache Front End configuration for web logic load balancing ....
Unfortunately all user session is on midTier, so if the midTier goes down, there goes their user session, any failover to another midTier will require them to relogin to the midTier. UNLESS you enable session broadcasting on all the Tomcat Servers, this means that when a user logs into that MidTier which creates a servlet session then that session is broadcast to all the other Tomcat servers. On Sep 21, 9:49 am, "Garrison, Sean (Norcross)" wrote: > Just curious if anyone has done any configuration of an apache front end web > server for a back end mid-tier server. Our server environment requires a DMZ > so we are fronting our mid-tiers through apache servers. What we were not > able to figure out was how to load balance the connection from the mid-tier > to the apache without getting . Our end goal is this: > > Load Balancer1 > / | \ > / | \ > Apache 1 Apache2 Apache3 > \ | / > \ | / > Load Balancer 2 > / | \ > / | \ > MT1 MT2 MT3 > \ | / > \ | / > Load Balancer 3 > | > | > ARS Server Group > > Currently we have this: > > Load Balancer1 > / | \ > / | \ > Apache 1 Apache2 Apache3 > | | | > MT1 MT2 MT3 > \ | / > \ | / > Load Balancer 3 > | > | > ARS Server Group > > The reason we wish to move to this configuration is so that if a mid-tier > goes down the user's session does not get lost. In our current > configuration if we bring a mid-tier down for maintenance a user's session > must be re-established. > > There are several apache modules that handle sessions for you. So we were > wondering if anyone has tried any apache modules and got it to work like > that. Mod_tsession, mod_session, and mod_wl are some of the modules that > come to mind that may help but we have not tried them. I remember BMC told > us that we must have a one to one with apache to make this work (this was > told to engineering and don't really know all the reasons). I was wondering > if there have been any new developments or if anyone has gotten it to work > this way even though it may not be "Supported" by BMC. > > Our environment: > > Apache 2 front end web servers > Mid-Tier 7.5 patch 6 running on weblogic 10 > ARS 7.5 patch 6 > Oracle 10g > > Thanks, > > Sean > > ___ > UNSUBSCRIBE or access ARSlist Archives atwww.arslist.org > attend wwrug11www.wwrug.comARSList: "Where the Answers Are" ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Apache Front End configuration for web logic load balancing ....
Just curious if anyone has done any configuration of an apache front end web server for a back end mid-tier server. Our server environment requires a DMZ so we are fronting our mid-tiers through apache servers. What we were not able to figure out was how to load balance the connection from the mid-tier to the apache without getting . Our end goal is this: Load Balancer1 / | \ / |\ Apache 1 Apache2Apache3 \ |/ \ | / Load Balancer 2 / | \ / |\ MT1 MT2 MT3 \ |/ \ | / Load Balancer 3 | | ARS Server Group Currently we have this: Load Balancer1 / | \ / |\ Apache 1 Apache2Apache3 | | | MT1 MT2 MT3 \ |/ \ | / Load Balancer 3 | | ARS Server Group The reason we wish to move to this configuration is so that if a mid-tier goes down the user's session does not get lost. In our current configuration if we bring a mid-tier down for maintenance a user's session must be re-established. There are several apache modules that handle sessions for you. So we were wondering if anyone has tried any apache modules and got it to work like that. Mod_tsession, mod_session, and mod_wl are some of the modules that come to mind that may help but we have not tried them. I remember BMC told us that we must have a one to one with apache to make this work (this was told to engineering and don't really know all the reasons). I was wondering if there have been any new developments or if anyone has gotten it to work this way even though it may not be "Supported" by BMC. Our environment: Apache 2 front end web servers Mid-Tier 7.5 patch 6 running on weblogic 10 ARS 7.5 patch 6 Oracle 10g Thanks, Sean ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.04 - mid tier load balancing settings
If you are using IIS in front of Tomcat serving as a JSP engine, you can enable logging on IIS to display these two fields which are not displayed on the logs by default.. sc-bytes ===> This is the number of bytes the server sends to the client. cs-bytes ===> This is the number of bytes the client sends to the server. These will give you an indication of exact packet sizes sent back and forth.. Packet sizes have to be incremented in blocks of 4K (4096 bytes). You cannot just increment it to a random number.. The defaults are 8K (8192 bytes) which is the lower limit of what you can set it to. There is also an upper limit to this size which is 64K (65536 bytes). This is what we learnt when we got bitten by that error.. For all practical reasons you never have to go that high.. For us we never got that error after we had it set to 12288 bytes even with queries that had a large qualification line that caused a sc-bytes packet to go over the 8192 mark.. Joe PS: If this hits your mailbox twice, its because I resent it as I got a mail back from the list serv rejecting my previous email saying it was over 1000 lines?? I get this error once in a blue moon even when its obvious its not over 1000 lines.. From: LJ LongWing Sent: Tuesday, April 26, 2011 11:55 AM Newsgroups: public.remedy.arsystem.general To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings ** Ours was here /conf/server.xml, search for maxHttpHeaderSize we just kept doubling the value until we stopped having errors…I know it’s not scientific, but it worked. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Joe Martin D'Souza Sent: Tuesday, April 26, 2011 9:46 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings ** We got bitten by that on my last engagement for a while :-) and it took us a while to find out why – after enabling the IIS cs-bytes and sc-bytes where we noticed the sc-bytes exceeded the defined default limit of 8192.. Tauf, It’s the max_packet_size you need to set on Tomcat in the worker.properties file. You will find enough info on how and where to set this if you google.. Joe ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.04 - mid tier load balancing settings
Ours was here /conf/server.xml, search for maxHttpHeaderSize we just kept doubling the value until we stopped having errors…I know it’s not scientific, but it worked. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Joe Martin D'Souza Sent: Tuesday, April 26, 2011 9:46 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings ** We got bitten by that on my last engagement for a while :-) and it took us a while to find out why – after enabling the IIS cs-bytes and sc-bytes where we noticed the sc-bytes exceeded the defined default limit of 8192.. Tauf, It’s the max_packet_size you need to set on Tomcat in the worker.properties file. You will find enough info on how and where to set this if you google.. Joe From: LJ LongWing <mailto:lj.longw...@gmail.com> Sent: Tuesday, April 26, 2011 11:26 AM Newsgroups: public.remedy.arsystem.general To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings ** I found that the unable to setup data connection errors (in the past) were related to the Tomcat MaxHeader size being too small….it defaults to something like 8K…with the SSO we are using our header size can get huge, so we have it set to something like 128. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Tuesday, April 26, 2011 9:14 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings ** Thanks LJ, have you noticed that your users are getting any RPC or errors that say “Unable to setup Data connection” errors? We are getting that and I’m not sure if this is due to the app load balancer or the front end web. Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Mobile:646.483.2779 _ From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of LJ LongWing Sent: Tuesday, April 26, 2011 10:54 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings Tauf, I truly wish I could give you some sage advice from a seasoned master, but I fear that you and I are groping around in the same darkness trying to put together a 1000 piece puzzle. I have the same docs that you have and am planning on assuming (yes, very dangerous, I know) that the ‘defaults’ are a good starting point, and make modifications as monitoring goes on as we reach production mid next month. Regarding the #’s you provided below….I think the settings you have below look reasonable. Obviously the default of unchecked lifespan and a lifespan of 0 won’t balance. Ironically we have also taken the Max Connections per server from the default of 80 to 200 as well because of problems we were experiencing with them being exhausted. I don’t see any reason I wouldn’t be able to use the ones you outlined below…ours is a 4 Mid-Tier/2 App server setup. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Tuesday, April 26, 2011 4:14 AM To: arslist@ARSLIST.ORG Subject: 7.6.04 - mid tier loadbalancing settings ** All (LJ), For those of you have gone through the load balancing exercise, can someone give a breakdown of the fields in the Connections section of the new mid tier? Yes, I’ve looked through the guide but it doesn’t clearly state what the ramifications of setting a number too high or too low means for something like, “Lifespan connection” or “Connection timeout.” Do these need to be greater or lower than what is set on the hardware load balancer? Has anyone found some settings that work better for them than others based on # of users or type of requests? Interestingly, BMC could not give me ANY recommendations or best practice guidance… so me and the network guys set the configs to what we thought made sense with a lot of assumptions on the settings below. The load balancer (between mid tier and app servers) is set to distribute based on Least Connection, with no persistency set. In my environment here are the settings: Enable Lifespan – Checked Connection Lifespan (minutes) = 60 Connection pool settings: Max connections per server = 200 Connection Timeout (minutes) = 30 Idle Connections per server = 5 Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Informatics-Infrastructure Office: 631.858.7765 Mobile:646.483.2779 _ This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to t
Re: 7.6.04 - mid tier load balancing settings
We got bitten by that on my last engagement for a while :-) and it took us a while to find out why – after enabling the IIS cs-bytes and sc-bytes where we noticed the sc-bytes exceeded the defined default limit of 8192.. Tauf, It’s the max_packet_size you need to set on Tomcat in the worker.properties file. You will find enough info on how and where to set this if you google.. Joe From: LJ LongWing Sent: Tuesday, April 26, 2011 11:26 AM Newsgroups: public.remedy.arsystem.general To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings ** I found that the unable to setup data connection errors (in the past) were related to the Tomcat MaxHeader size being too small….it defaults to something like 8K…with the SSO we are using our header size can get huge, so we have it set to something like 128. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Tuesday, April 26, 2011 9:14 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings ** Thanks LJ, have you noticed that your users are getting any RPC or errors that say “Unable to setup Data connection” errors? We are getting that and I’m not sure if this is due to the app load balancer or the front end web. Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Mobile:646.483.2779 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of LJ LongWing Sent: Tuesday, April 26, 2011 10:54 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings Tauf, I truly wish I could give you some sage advice from a seasoned master, but I fear that you and I are groping around in the same darkness trying to put together a 1000 piece puzzle. I have the same docs that you have and am planning on assuming (yes, very dangerous, I know) that the ‘defaults’ are a good starting point, and make modifications as monitoring goes on as we reach production mid next month. Regarding the #’s you provided below….I think the settings you have below look reasonable. Obviously the default of unchecked lifespan and a lifespan of 0 won’t balance. Ironically we have also taken the Max Connections per server from the default of 80 to 200 as well because of problems we were experiencing with them being exhausted. I don’t see any reason I wouldn’t be able to use the ones you outlined below…ours is a 4 Mid-Tier/2 App server setup. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Tuesday, April 26, 2011 4:14 AM To: arslist@ARSLIST.ORG Subject: 7.6.04 - mid tier loadbalancing settings ** All (LJ), For those of you have gone through the load balancing exercise, can someone give a breakdown of the fields in the Connections section of the new mid tier? Yes, I’ve looked through the guide but it doesn’t clearly state what the ramifications of setting a number too high or too low means for something like, “Lifespan connection” or “Connection timeout.” Do these need to be greater or lower than what is set on the hardware load balancer? Has anyone found some settings that work better for them than others based on # of users or type of requests? Interestingly, BMC could not give me ANY recommendations or best practice guidance… so me and the network guys set the configs to what we thought made sense with a lot of assumptions on the settings below. The load balancer (between mid tier and app servers) is set to distribute based on Least Connection, with no persistency set. In my environment here are the settings: Enable Lifespan – Checked Connection Lifespan (minutes) = 60 Connection pool settings: Max connections per server = 200 Connection Timeout (minutes) = 30 Idle Connections per server = 5 Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Informatics-Infrastructure Office: 631.858.7765 Mobile:646.483.2779 This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of
Re: 7.6.04 - mid tier load balancing settings
Can you point me to where to set that? I'm a noob when it comes to tomcat... Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Mobile:646.483.2779 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of LJ LongWing Sent: Tuesday, April 26, 2011 11:26 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings I found that the unable to setup data connection errors (in the past) were related to the Tomcat MaxHeader size being too smallit defaults to something like 8K...with the SSO we are using our header size can get huge, so we have it set to something like 128. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Tuesday, April 26, 2011 9:14 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings ** Thanks LJ, have you noticed that your users are getting any RPC or errors that say "Unable to setup Data connection" errors? We are getting that and I'm not sure if this is due to the app load balancer or the front end web. Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Mobile:646.483.2779 _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ ** This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.04 - mid tier load balancing settings
I found that the unable to setup data connection errors (in the past) were related to the Tomcat MaxHeader size being too small..it defaults to something like 8K.with the SSO we are using our header size can get huge, so we have it set to something like 128. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Tuesday, April 26, 2011 9:14 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings ** Thanks LJ, have you noticed that your users are getting any RPC or errors that say "Unable to setup Data connection" errors? We are getting that and I'm not sure if this is due to the app load balancer or the front end web. Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Mobile:646.483.2779 _ From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of LJ LongWing Sent: Tuesday, April 26, 2011 10:54 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings Tauf, I truly wish I could give you some sage advice from a seasoned master, but I fear that you and I are groping around in the same darkness trying to put together a 1000 piece puzzle. I have the same docs that you have and am planning on assuming (yes, very dangerous, I know) that the 'defaults' are a good starting point, and make modifications as monitoring goes on as we reach production mid next month. Regarding the #'s you provided below..I think the settings you have below look reasonable. Obviously the default of unchecked lifespan and a lifespan of 0 won't balance. Ironically we have also taken the Max Connections per server from the default of 80 to 200 as well because of problems we were experiencing with them being exhausted. I don't see any reason I wouldn't be able to use the ones you outlined below.ours is a 4 Mid-Tier/2 App server setup. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Tuesday, April 26, 2011 4:14 AM To: arslist@ARSLIST.ORG Subject: 7.6.04 - mid tier loadbalancing settings ** All (LJ), For those of you have gone through the load balancing exercise, can someone give a breakdown of the fields in the Connections section of the new mid tier? Yes, I've looked through the guide but it doesn't clearly state what the ramifications of setting a number too high or too low means for something like, "Lifespan connection" or "Connection timeout." Do these need to be greater or lower than what is set on the hardware load balancer? Has anyone found some settings that work better for them than others based on # of users or type of requests? Interestingly, BMC could not give me ANY recommendations or best practice guidance. so me and the network guys set the configs to what we thought made sense with a lot of assumptions on the settings below. The load balancer (between mid tier and app servers) is set to distribute based on Least Connection, with no persistency set. In my environment here are the settings: Enable Lifespan - Checked Connection Lifespan (minutes) = 60 Connection pool settings: Max connections per server = 200 Connection Timeout (minutes) = 30 Idle Connections per server = 5 Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Informatics-Infrastructure Office: 631.858.7765 Mobile:646.483.2779 _ This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ _ This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments
Re: 7.6.04 - mid tier load balancing settings
Thanks LJ, have you noticed that your users are getting any RPC or errors that say "Unable to setup Data connection" errors? We are getting that and I'm not sure if this is due to the app load balancer or the front end web. Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Mobile:646.483.2779 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of LJ LongWing Sent: Tuesday, April 26, 2011 10:54 AM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - mid tier load balancing settings Tauf, I truly wish I could give you some sage advice from a seasoned master, but I fear that you and I are groping around in the same darkness trying to put together a 1000 piece puzzle. I have the same docs that you have and am planning on assuming (yes, very dangerous, I know) that the 'defaults' are a good starting point, and make modifications as monitoring goes on as we reach production mid next month. Regarding the #'s you provided belowI think the settings you have below look reasonable. Obviously the default of unchecked lifespan and a lifespan of 0 won't balance. Ironically we have also taken the Max Connections per server from the default of 80 to 200 as well because of problems we were experiencing with them being exhausted. I don't see any reason I wouldn't be able to use the ones you outlined below...ours is a 4 Mid-Tier/2 App server setup. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Tuesday, April 26, 2011 4:14 AM To: arslist@ARSLIST.ORG Subject: 7.6.04 - mid tier loadbalancing settings ** All (LJ), For those of you have gone through the load balancing exercise, can someone give a breakdown of the fields in the Connections section of the new mid tier? Yes, I've looked through the guide but it doesn't clearly state what the ramifications of setting a number too high or too low means for something like, "Lifespan connection" or "Connection timeout." Do these need to be greater or lower than what is set on the hardware load balancer? Has anyone found some settings that work better for them than others based on # of users or type of requests? Interestingly, BMC could not give me ANY recommendations or best practice guidance... so me and the network guys set the configs to what we thought made sense with a lot of assumptions on the settings below. The load balancer (between mid tier and app servers) is set to distribute based on Least Connection, with no persistency set. In my environment here are the settings: Enable Lifespan - Checked Connection Lifespan (minutes) = 60 Connection pool settings: Max connections per server = 200 Connection Timeout (minutes) = 30 Idle Connections per server = 5 Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Informatics-Infrastructure Office: 631.858.7765 Mobile:646.483.2779 This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ ** This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. ___
Re: 7.6.04 - mid tier load balancing settings
Tauf, I truly wish I could give you some sage advice from a seasoned master, but I fear that you and I are groping around in the same darkness trying to put together a 1000 piece puzzle. I have the same docs that you have and am planning on assuming (yes, very dangerous, I know) that the 'defaults' are a good starting point, and make modifications as monitoring goes on as we reach production mid next month. Regarding the #'s you provided below..I think the settings you have below look reasonable. Obviously the default of unchecked lifespan and a lifespan of 0 won't balance. Ironically we have also taken the Max Connections per server from the default of 80 to 200 as well because of problems we were experiencing with them being exhausted. I don't see any reason I wouldn't be able to use the ones you outlined below.ours is a 4 Mid-Tier/2 App server setup. From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Tuesday, April 26, 2011 4:14 AM To: arslist@ARSLIST.ORG Subject: 7.6.04 - mid tier loadbalancing settings ** All (LJ), For those of you have gone through the load balancing exercise, can someone give a breakdown of the fields in the Connections section of the new mid tier? Yes, I've looked through the guide but it doesn't clearly state what the ramifications of setting a number too high or too low means for something like, "Lifespan connection" or "Connection timeout." Do these need to be greater or lower than what is set on the hardware load balancer? Has anyone found some settings that work better for them than others based on # of users or type of requests? Interestingly, BMC could not give me ANY recommendations or best practice guidance. so me and the network guys set the configs to what we thought made sense with a lot of assumptions on the settings below. The load balancer (between mid tier and app servers) is set to distribute based on Least Connection, with no persistency set. In my environment here are the settings: Enable Lifespan - Checked Connection Lifespan (minutes) = 60 Connection pool settings: Max connections per server = 200 Connection Timeout (minutes) = 30 Idle Connections per server = 5 Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Informatics-Infrastructure Office: 631.858.7765 Mobile:646.483.2779 _ This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
7.6.04 - Load balancing
Axton I believe the problem is the documentation strongly suggests it's possible, and the reality is that Midtier probably doesn't. The good news is that SSO Plugin 3.3.6 is distributable! John ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.04 - Load balancing
I think there may be more that just session information that needs to be serialized in the case of the mid-tier. The mid-tier sticks things out on the filesystem that it expects subsequent requests to be able to access (thinking of the report work directory and attachments). Axton On Fri, Apr 22, 2011 at 2:45 PM, John Baker wrote: > LJ, > > That does suggest it should work without the sticky bit, but the web.xml > file is missing the element and hence Midtier won't (or > rather, should not) distribute sessions. > > > John > > -- > Single Sign On for AR System > http://www.javasystemsolutions.com/jss/ssoplugin > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.04 - Load balancing
When I see the term 'fail-over enabled', it says to me that you have multiple servers in a farm, of which 1 (and only 1) is active at a given point in time. If this is the case, then the sticky bit is not required as all requests will go to the same web server/servlet container. I do not think the mid-tier supports clustering on the J2EE side. Axton On Fri, Apr 22, 2011 at 2:24 PM, LJ LongWing wrote: > John, > On page 19, bout half way down it states > > Load balancing between the clients and web servers without setting a sticky > bit > > If the web servers in your AR System were installed in a cluster (with > fail-over enabled), then setting the sticky bit on the load balancer > between > the clients and web servers is not needed. > > This tells me directly that I can turn the sticky bit off on my web load > balancer if I cluster my web server. > > Am I reading that wrong?yes, I may need to add the distributable > attribute to the mid-tier web.xml, but that shows as 'ok' to me based on > the > above statement from BMC, right? > > -Original Message- > From: Action Request System discussion list(ARSList) > [mailto:arslist@ARSLIST.ORG] On Behalf Of John Baker > Sent: Friday, April 22, 2011 12:16 PM > To: arslist@ARSLIST.ORG > Subject: 7.6.04 - Load balancing > > Hello, > > I'm referring to the load balancing document and I think there may be some > confusion. > > The document states the sticky bit isn't required for the LB between > MIdtier > and AR System server, but I can not see (and forgive me for not fully > reading the document in detail) any reference to the LB in front of the > Midtiers and the sticky bit. I therefore conclude that the LB in front of > the Midtiers needs the sticky bit set because the Midtiers do not cluster. > > Further to that, the Midtier web.xml file doesn't contain: > > > > Which as you can see in this page: > > http://tomcat.apache.org/tomcat-6.0-doc/cluster-howto.html > > (Under clustering basics) > > needs to be included for Tomcat (and other servlet engines) to distribute > sessions. > > > John > > -- > Single Sign On for AR System > http://www.javasystemsolutions.com/jss/ssoplugin > > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are" > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are" > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.04 - Load balancing
Does 7.6 support clustering on the J2EE side (serialization of session information, uploads, reports, etc.)? Previous releases of the mid-tier required that some type of session persistence be in use (source ip, cookie, jsessionid, etc.). If 7.6 still requires session persistence for the mid-tier and you are using f5 LTM devices, ensure that you have a OneConnect profile associated with the VIP and you are using some form of session persistence. This will resolve the session timeout/invalid session errors. With the 7.6 arserver, I've read that it is no longer necessary to use session persistence to the arservers. I don't remember reading that the same applies to the mid-tier. Axton On Fri, Apr 22, 2011 at 11:21 AM, LJ LongWing wrote: > ** > > Tauf, > > I’m in the middle of setting this exact config up. Two load balancers, one > in front of the web, another between web and app. The only issues we are > experiencing are related to session timeout problems, which is server > clustering related…still working on that one. > > > > If you are getting plugin errors, ensure that in your armonitor.cfg file > you are starting your java plugin server connecting to your individual node > alias, and in the ar.cfg file, ensure that all of your java plugins are > connecting to the same node based name… > > > > If you are by chance using Tomcat for your clustering, I would be willing > to accept config tips/files to help smooth my road J > > > > *From:* Action Request System discussion list(ARSList) [mailto: > arslist@ARSLIST.ORG] *On Behalf Of *Chowdhury, Tauf > *Sent:* Friday, April 22, 2011 9:31 AM > *To:* arslist@ARSLIST.ORG > *Subject:* 7.6.04 - Load balancing > > > > ** > > Has anyone set up Remedy 7.6.04 as a server group with a hardware load > balancer in between the mid tier and AR servers? > > Has that gone well? > > We set the “enable lifespan” option on the mid tiers and had the load > balancer clear the persistency flag and are seeing odd plugin errors. Not > sure if it’s related. I have a ticket open with BMC but curious if anyone > experienced the issue. > > > > *T**auf** **C**howdhury** **|** **F**orest** **L**aboratories**, **I**nc.* > ** > > Analyst, Service Management > > Informatics-Infrastructure > > Office: 631.858.7765 > > Mobile:646.483.2779 > > > > > -- > > This e-mail and its attachments may contain Forest Laboratories, Inc. > proprietary information that is privileged, confidential or subject to > copyright belonging to Forest Laboratories, Inc. This e-mail is intended > solely for the use of the individual or entity to which it is addressed. If > you are not the intended recipient of this e-mail, or the employee or agent > responsible for delivering this e-mail to the intended recipient, you are > hereby notified that any dissemination, distribution, copying or action > taken in relation to the contents of and attachments to this e-mail is > strictly prohibited and may be unlawful. If you have received this e-mail in > error, please notify the sender immediately and permanently delete the > original and any copy of this e-mail and any printout. > > _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ > _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
7.6.04 - Load balancing
LJ, If you want to look into clustering, start in the Tomcat server.xml file: John -- Single Sign On for AR System http://www.javasystemsolutions.com/jss/ssoplugin ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
7.6.04 - Load balancing
LJ, That does suggest it should work without the sticky bit, but the web.xml file is missing the element and hence Midtier won't (or rather, should not) distribute sessions. John -- Single Sign On for AR System http://www.javasystemsolutions.com/jss/ssoplugin ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.04 - Load balancing
John, On page 19, bout half way down it states Load balancing between the clients and web servers without setting a sticky bit If the web servers in your AR System were installed in a cluster (with fail-over enabled), then setting the sticky bit on the load balancer between the clients and web servers is not needed. This tells me directly that I can turn the sticky bit off on my web load balancer if I cluster my web server. Am I reading that wrong?yes, I may need to add the distributable attribute to the mid-tier web.xml, but that shows as 'ok' to me based on the above statement from BMC, right? -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of John Baker Sent: Friday, April 22, 2011 12:16 PM To: arslist@ARSLIST.ORG Subject: 7.6.04 - Load balancing Hello, I'm referring to the load balancing document and I think there may be some confusion. The document states the sticky bit isn't required for the LB between MIdtier and AR System server, but I can not see (and forgive me for not fully reading the document in detail) any reference to the LB in front of the Midtiers and the sticky bit. I therefore conclude that the LB in front of the Midtiers needs the sticky bit set because the Midtiers do not cluster. Further to that, the Midtier web.xml file doesn't contain: Which as you can see in this page: http://tomcat.apache.org/tomcat-6.0-doc/cluster-howto.html (Under clustering basics) needs to be included for Tomcat (and other servlet engines) to distribute sessions. John -- Single Sign On for AR System http://www.javasystemsolutions.com/jss/ssoplugin ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are" ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.04 - Load balancing
Have you tried setting your plugin logs to fine? What java are you using? 32 bit? 64 bit? Are you mixing between the two? Meaning 32 bit for some components and 64 for others? If so eliminate that if possible and use either 32 or 64 across the board.. Joe From: Chowdhury, Tauf Sent: Friday, April 22, 2011 1:14 PM Newsgroups: public.remedy.arsystem.general To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - Load balancing ** Joe, One example off odd behavior is in SRM, if I try to view anything that contains a data viz field like the PDT or Process Tab on an open request, it’s either a blank screen or it says “error loading document” Both of these fields have AL’s that trigger which perform Run process functions. Also, when I submit a service request, it’s not creating the incident. I ran filter/api/sql logging to no avail. I was getting an error stating Unable to connect to Ar System Plugin Server when I tried submitting an incident (after marking yes to the option in Rules that creates request on submit). A filter (don’t have the name right now) like SRM USM… failed and it looked like it was doing a run process as well. It may not be related to load balancing at all and a plugin configuration error but I am lost right now… Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Mobile:646.483.2779 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Joe Martin D'Souza Sent: Friday, April 22, 2011 12:31 PM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - Load balancing Define ‘odd plugin errors’ Tauf... I have seen plugin errors with 7.6.03 that seem to be resolved in 7.6.04 that are not related to load balancing but I suspect is related to the fact that 7.6.03 was not yet supporting 64 bit java at all points wherever Java was required.. On 7.6.03 for eg, if you used 64 bit java wherever 64 bit is supported and 32 wherever 64 bit was not (which is exactly what I had done) I was unable to set the Server Ports and Queues using the plugin interface. Changing values there would not result in any errors, not even in the plugin logs when set to fine, but it just wouldn’t change the configuration. Close that administration form and reopen it and those values would read the current values (unchanged) from the config file, and your new values would not be seen. Even from the timestamp on the ar.conf file it was clear that it was not even touching the file.. Remedy support had recognized it as a bug.. It was resolved when I used 64 bit java all round in 7.6.04 where areas where 64 bit java was not supported was then supported.. Which is why I get that feeling it was because java versions were mixed and matched in 7.6.03, it may have created that problem.. What is the nature of your problem? Joe From: LJ LongWing Sent: Friday, April 22, 2011 12:21 PM Newsgroups: public.remedy.arsystem.general To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - Load balancing ** Tauf, I’m in the middle of setting this exact config up. Two load balancers, one in front of the web, another between web and app. The only issues we are experiencing are related to session timeout problems, which is server clustering related…still working on that one. If you are getting plugin errors, ensure that in your armonitor.cfg file you are starting your java plugin server connecting to your individual node alias, and in the ar.cfg file, ensure that all of your java plugins are connecting to the same node based name… If you are by chance using Tomcat for your clustering, I would be willing to accept config tips/files to help smooth my road J From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Friday, April 22, 2011 9:31 AM To: arslist@ARSLIST.ORG Subject: 7.6.04 - Load balancing ** Has anyone set up Remedy 7.6.04 as a server group with a hardware load balancer in between the mid tier and AR servers? Has that gone well? We set the “enable lifespan” option on the mid tiers and had the load balancer clear the persistency flag and are seeing odd plugin errors. Not sure if it’s related. I have a ticket open with BMC but curious if anyone experienced the issue. Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Informatics-Infrastructure Office: 631.858.7765 Mobile:646.483.2779 ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
7.6.04 - Load balancing
Hello, I'm referring to the load balancing document and I think there may be some confusion. The document states the sticky bit isn't required for the LB between MIdtier and AR System server, but I can not see (and forgive me for not fully reading the document in detail) any reference to the LB in front of the Midtiers and the sticky bit. I therefore conclude that the LB in front of the Midtiers needs the sticky bit set because the Midtiers do not cluster. Further to that, the Midtier web.xml file doesn't contain: Which as you can see in this page: http://tomcat.apache.org/tomcat-6.0-doc/cluster-howto.html (Under clustering basics) needs to be included for Tomcat (and other servlet engines) to distribute sessions. John -- Single Sign On for AR System http://www.javasystemsolutions.com/jss/ssoplugin ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.04 - Load balancing
Joe, One example off odd behavior is in SRM, if I try to view anything that contains a data viz field like the PDT or Process Tab on an open request, it's either a blank screen or it says "error loading document" Both of these fields have AL's that trigger which perform Run process functions. Also, when I submit a service request, it's not creating the incident. I ran filter/api/sql logging to no avail. I was getting an error stating Unable to connect to Ar System Plugin Server when I tried submitting an incident (after marking yes to the option in Rules that creates request on submit). A filter (don't have the name right now) like SRM USM... failed and it looked like it was doing a run process as well. It may not be related to load balancing at all and a plugin configuration error but I am lost right now... Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Mobile:646.483.2779 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Joe Martin D'Souza Sent: Friday, April 22, 2011 12:31 PM To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - Load balancing Define 'odd plugin errors' Tauf... I have seen plugin errors with 7.6.03 that seem to be resolved in 7.6.04 that are not related to load balancing but I suspect is related to the fact that 7.6.03 was not yet supporting 64 bit java at all points wherever Java was required.. On 7.6.03 for eg, if you used 64 bit java wherever 64 bit is supported and 32 wherever 64 bit was not (which is exactly what I had done) I was unable to set the Server Ports and Queues using the plugin interface. Changing values there would not result in any errors, not even in the plugin logs when set to fine, but it just wouldn't change the configuration. Close that administration form and reopen it and those values would read the current values (unchanged) from the config file, and your new values would not be seen. Even from the timestamp on the ar.conf file it was clear that it was not even touching the file.. Remedy support had recognized it as a bug.. It was resolved when I used 64 bit java all round in 7.6.04 where areas where 64 bit java was not supported was then supported.. Which is why I get that feeling it was because java versions were mixed and matched in 7.6.03, it may have created that problem.. What is the nature of your problem? Joe From: LJ LongWing <mailto:lj.longw...@gmail.com> Sent: Friday, April 22, 2011 12:21 PM Newsgroups: public.remedy.arsystem.general To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - Load balancing ** Tauf, I'm in the middle of setting this exact config up. Two load balancers, one in front of the web, another between web and app. The only issues we are experiencing are related to session timeout problems, which is server clustering related...still working on that one. If you are getting plugin errors, ensure that in your armonitor.cfg file you are starting your java plugin server connecting to your individual node alias, and in the ar.cfg file, ensure that all of your java plugins are connecting to the same node based name... If you are by chance using Tomcat for your clustering, I would be willing to accept config tips/files to help smooth my road :-) From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Friday, April 22, 2011 9:31 AM To: arslist@ARSLIST.ORG Subject: 7.6.04 - Load balancing ** Has anyone set up Remedy 7.6.04 as a server group with a hardware load balancer in between the mid tier and AR servers? Has that gone well? We set the "enable lifespan" option on the mid tiers and had the load balancer clear the persistency flag and are seeing odd plugin errors. Not sure if it's related. I have a ticket open with BMC but curious if anyone experienced the issue. Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Informatics-Infrastructure Office: 631.858.7765 Mobile:646.483.2779 This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. _attend WWRUG11 www
Re: 7.6.04 - Load balancing
Define ‘odd plugin errors’ Tauf... I have seen plugin errors with 7.6.03 that seem to be resolved in 7.6.04 that are not related to load balancing but I suspect is related to the fact that 7.6.03 was not yet supporting 64 bit java at all points wherever Java was required.. On 7.6.03 for eg, if you used 64 bit java wherever 64 bit is supported and 32 wherever 64 bit was not (which is exactly what I had done) I was unable to set the Server Ports and Queues using the plugin interface. Changing values there would not result in any errors, not even in the plugin logs when set to fine, but it just wouldn’t change the configuration. Close that administration form and reopen it and those values would read the current values (unchanged) from the config file, and your new values would not be seen. Even from the timestamp on the ar.conf file it was clear that it was not even touching the file.. Remedy support had recognized it as a bug.. It was resolved when I used 64 bit java all round in 7.6.04 where areas where 64 bit java was not supported was then supported.. Which is why I get that feeling it was because java versions were mixed and matched in 7.6.03, it may have created that problem.. What is the nature of your problem? Joe From: LJ LongWing Sent: Friday, April 22, 2011 12:21 PM Newsgroups: public.remedy.arsystem.general To: arslist@ARSLIST.ORG Subject: Re: 7.6.04 - Load balancing ** Tauf, I’m in the middle of setting this exact config up. Two load balancers, one in front of the web, another between web and app. The only issues we are experiencing are related to session timeout problems, which is server clustering related…still working on that one. If you are getting plugin errors, ensure that in your armonitor.cfg file you are starting your java plugin server connecting to your individual node alias, and in the ar.cfg file, ensure that all of your java plugins are connecting to the same node based name… If you are by chance using Tomcat for your clustering, I would be willing to accept config tips/files to help smooth my road J From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Friday, April 22, 2011 9:31 AM To: arslist@ARSLIST.ORG Subject: 7.6.04 - Load balancing ** Has anyone set up Remedy 7.6.04 as a server group with a hardware load balancer in between the mid tier and AR servers? Has that gone well? We set the “enable lifespan” option on the mid tiers and had the load balancer clear the persistency flag and are seeing odd plugin errors. Not sure if it’s related. I have a ticket open with BMC but curious if anyone experienced the issue. Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Informatics-Infrastructure Office: 631.858.7765 Mobile:646.483.2779 This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.04 - Load balancing
Tauf, I'm in the middle of setting this exact config up. Two load balancers, one in front of the web, another between web and app. The only issues we are experiencing are related to session timeout problems, which is server clustering related.still working on that one. If you are getting plugin errors, ensure that in your armonitor.cfg file you are starting your java plugin server connecting to your individual node alias, and in the ar.cfg file, ensure that all of your java plugins are connecting to the same node based name. If you are by chance using Tomcat for your clustering, I would be willing to accept config tips/files to help smooth my road J From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Chowdhury, Tauf Sent: Friday, April 22, 2011 9:31 AM To: arslist@ARSLIST.ORG Subject: 7.6.04 - Load balancing ** Has anyone set up Remedy 7.6.04 as a server group with a hardware load balancer in between the mid tier and AR servers? Has that gone well? We set the "enable lifespan" option on the mid tiers and had the load balancer clear the persistency flag and are seeing odd plugin errors. Not sure if it's related. I have a ticket open with BMC but curious if anyone experienced the issue. Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Informatics-Infrastructure Office: 631.858.7765 Mobile:646.483.2779 _ This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
7.6.04 - Load balancing
Has anyone set up Remedy 7.6.04 as a server group with a hardware load balancer in between the mid tier and AR servers? Has that gone well? We set the "enable lifespan" option on the mid tiers and had the load balancer clear the persistency flag and are seeing odd plugin errors. Not sure if it's related. I have a ticket open with BMC but curious if anyone experienced the issue. Tauf Chowdhury | Forest Laboratories, Inc. Analyst, Service Management Informatics-Infrastructure Office: 631.858.7765 Mobile:646.483.2779 ** This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.3 Load Balancing AR Server / Mid Tier
I have more info... The source ports are limited on our proxy to 10,000. The users are showing up quickly on the firewall / loadbalancer, and Checkpoint firewall is questioning the usage - thinking this is an intruder of some sort. The firewall is not creating a new connection. The source port stays open. As each user accesses, it looks like 1 IP address. Each user gets 1 port. Example: User "Joe" logins in on port 123, then signs off (the connection is staying open). The connection shows active. The next user "Bill" logs in on the same source port 123. The same source port remains open in the Connection Table. For some reason when Joe logged off, the port 123 did not close in the Connection Table. This did not happen in 7.1, however we are seeing this issue in 7.6 In a message dated 11/17/2010 12:19:39 P.M. Eastern Standard Time, dave.shell...@tycoelectronics.com writes: ** Kathy, There are a couple of potential issues with load balancer. The first I can think of is with the "sticky bit" which sends the user to the same AR System and prevents the logging in from a different IP address issue. If you give a little more detail about the issue, those with more load balancing expertise may be able to help you. Dave From: Action Request System discussion list(ARSList) [mailto:arsl...@arslist.org] On Behalf Of Kathy Morris Sent: Wednesday, November 17, 2010 1:15 PM To: arslist@ARSLIST.ORG Subject: 7.6.3 Load Balancing AR Server / Mid Tier ** Hi, We are experiencing serious connection issues with 7.6.03 AR System / ITSM and Cisco. Apparently there is an issue with Load Balancing. They are using Cisco as a load balancer. Is there any documentation for load balancing with 7.6.03? I saw documents for 7.1 however we were looking for more recent info. What types of issues could cause connection issues between the load balancer and the AR Server? I thought I heard something on the list a couple of months ago about this, however I have not located the email. _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.3 Load Balancing AR Server / Mid Tier
It sounds like there are 3 virtual gateways, each one has a pool of ports. The ports are being reused, and the sessions are remaining open. Is the # of ports on the proxy important? I would think so, however one of the team members responsible for the switch says the number of ports on the proxy is NOT important. There is mention of disabling the secure XL command on the firewall. To use encryption, port #443 must be open. In a message dated 11/17/2010 12:19:39 P.M. Eastern Standard Time, dave.shell...@tycoelectronics.com writes: ** Kathy, There are a couple of potential issues with load balancer. The first I can think of is with the "sticky bit" which sends the user to the same AR System and prevents the logging in from a different IP address issue. If you give a little more detail about the issue, those with more load balancing expertise may be able to help you. Dave From: Action Request System discussion list(ARSList) [mailto:arsl...@arslist.org] On Behalf Of Kathy Morris Sent: Wednesday, November 17, 2010 1:15 PM To: arslist@ARSLIST.ORG Subject: 7.6.3 Load Balancing AR Server / Mid Tier ** Hi, We are experiencing serious connection issues with 7.6.03 AR System / ITSM and Cisco. Apparently there is an issue with Load Balancing. They are using Cisco as a load balancer. Is there any documentation for load balancing with 7.6.03? I saw documents for 7.1 however we were looking for more recent info. What types of issues could cause connection issues between the load balancer and the AR Server? I thought I heard something on the list a couple of months ago about this, however I have not located the email. _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.3 Load Balancing AR Server / Mid Tier
Apparently the packets are being dropped from the Cisco switch. And connections are being used and reused. In a message dated 11/17/2010 12:19:39 P.M. Eastern Standard Time, dave.shell...@tycoelectronics.com writes: ** Kathy, There are a couple of potential issues with load balancer. The first I can think of is with the "sticky bit" which sends the user to the same AR System and prevents the logging in from a different IP address issue. If you give a little more detail about the issue, those with more load balancing expertise may be able to help you. Dave From: Action Request System discussion list(ARSList) [mailto:arsl...@arslist.org] On Behalf Of Kathy Morris Sent: Wednesday, November 17, 2010 1:15 PM To: arslist@ARSLIST.ORG Subject: 7.6.3 Load Balancing AR Server / Mid Tier ** Hi, We are experiencing serious connection issues with 7.6.03 AR System / ITSM and Cisco. Apparently there is an issue with Load Balancing. They are using Cisco as a load balancer. Is there any documentation for load balancing with 7.6.03? I saw documents for 7.1 however we were looking for more recent info. What types of issues could cause connection issues between the load balancer and the AR Server? I thought I heard something on the list a couple of months ago about this, however I have not located the email. _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: 7.6.3 Load Balancing AR Server / Mid Tier
Kathy, There are a couple of potential issues with load balancer. The first I can think of is with the "sticky bit" which sends the user to the same AR System and prevents the logging in from a different IP address issue. If you give a little more detail about the issue, those with more load balancing expertise may be able to help you. Dave From: Action Request System discussion list(ARSList) [mailto:arsl...@arslist.org] On Behalf Of Kathy Morris Sent: Wednesday, November 17, 2010 1:15 PM To: arslist@ARSLIST.ORG Subject: 7.6.3 Load Balancing AR Server / Mid Tier ** Hi, We are experiencing serious connection issues with 7.6.03 AR System / ITSM and Cisco. Apparently there is an issue with Load Balancing. They are using Cisco as a load balancer. Is there any documentation for load balancing with 7.6.03? I saw documents for 7.1 however we were looking for more recent info. What types of issues could cause connection issues between the load balancer and the AR Server? I thought I heard something on the list a couple of months ago about this, however I have not located the email. _attend WWRUG11 www.wwrug.com ARSlist: "Where the Answers Are"_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
7.6.3 Load Balancing AR Server / Mid Tier
Hi, We are experiencing serious connection issues with 7.6.03 AR System / ITSM and Cisco. Apparently there is an issue with Load Balancing. They are using Cisco as a load balancer. Is there any documentation for load balancing with 7.6.03? I saw documents for 7.1 however we were looking for more recent info. What types of issues could cause connection issues between the load balancer and the AR Server? I thought I heard something on the list a couple of months ago about this, however I have not located the email. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug11 www.wwrug.com ARSList: "Where the Answers Are"
Re: ARServer group and the ports that require load balancing
I concur - we are running mutliple everything with load balancers at my current customer and we had this same problem before we added the aliases. Interestingly we have several duplicate boxes (dev, cat, etc) and only 1 of them had this problem. I never did figure out why. We are not using port mapper of course. William Rentfrow Principal Consultant, StrataCom Inc. wrentf...@stratacominc.com 715-410-8056 C 715-592-5185 O From: Action Request System discussion list(ARSList) [mailto:arsl...@arslist.org] On Behalf Of Conny Martin Sent: Tuesday, April 07, 2009 3:51 AM To: arslist@ARSLIST.ORG Subject: AW: ARServer group and the ports that require load balancing George, IMHO this is the relevant part of the documentation. Specifying plug-in server aliases (Step 4e) When you specify a common AR System server alias, AR System uses the alias name to connect to the plug-in server. When the AR System server is connecting to its local plug-in server, the connection will not be successful if the server name resolves to anything other than the local server. To route connections successfully in an environment where the server name resolves to a load balancer, you must define plug-in server aliases. The purpose of this step is to designate the local server as the host, not to change the name of the plug-in. To specify plug-in server aliases Add an alias definition for each loaded plug-in to the configuration file, using the plug-in name specified in the plug-in definition. Use the following format, substituting the alias name, plug-in name, host name, and optional port number: Server-Plugin-Alias: [:] The AR System server connects to the plug-in server using host at port when the AR System server communicates with the plug-in. Host must resolve to the local AR System server, and the port number is specified if the plug-in server is listening on a particular port. The following definition is a typical example: Server-Plugin-Alias: ARSYS.ARDBC.REPORT ARSYS.ARDBC.REPORT myserver:2020 For more information about Server-Plugin-Alias, see “ar.conf (ar.cfg)” on page 248. We have a similar setup and solved this problem in a different way. We added the following line to /etc/hosts 127.0.0.1 itsm.telekom.de itsm.telekom.de is configured as common server alias in ar.conf which is used to connect to the plugin server. HTH Kind Regards Conny Von: Action Request System discussion list(ARSList) [mailto:arsl...@arslist.org] Im Auftrag von George Jungle Gesendet: Dienstag, 7. April 2009 03:21 An: arslist@ARSLIST.ORG Betreff: ARServer group and the ports that require load balancing ** Guys, Wondering if you can help me out. We've got an issue with the Plug-in server running within a high available environment of Remedy (2 Midtiers, 2 AR Servers, 2 DB Servers) with each tier separated by a firewall. There is one load balancer in front of the midtier servers and another in front of the AR Servers. So, basically the AR Servers are running within a Server Group. When we try to open up a Remedy form, such as the "Server Information" form, we get an error saying : ARERR [8939] The AR System Plug-In server is not responding. Cannot connect to the system at this time. Contact your AR System Administrator for assistance. : RPC: Timed out Currently the only port that is load balanced is the TCP port that the users access the AR Servers by. Should we also need to load balance the Plug-In port? (9000)? All help is appreciated. Below are some emails I've dug up just from searching the web... Thanks. George Re: RPC: Unable to receive Tony Worthington Mon, 06 Aug 2007 11:27:16 -0700 You probably need to review the BMC documents about plugins behind a VIP or firewall without a portmapper. "Using a Hardware Load Balancer with BMC Remedy Action Request System 7.0" page 9, and others that I can't find at the moment. Here's an example of some lines from one of our ar.conf's (this one from a server group, but relevant since we don't use portmapper either) Plugin-Port: 9000 Register-With-Portmapper: F TCD-Specific-Port: 8639 Server-Plugin-Alias: ARSYS.ARDBC.CONFIGFILE ARSYS.ARDBC.CONFIGFILE ksms562:9000 Server-Plugin-Alias: ARSYS.ARDBC.LDAP ARSYS.ARDBC.LDAP ksms562:9000 Server-Plugin-Alias: ARSYS.ARDBC.PREVIEW ARSYS.ARDBC.PREVIEW ksms562:9000 Server-Plugin-Alias: ARSYS.AREA.LDAP ARSYS.AREA.LDAP ksms562:9000 Server-Plugin-Alias: ARSYS.ARF.WEBSERVICE ARSYS.ARF.WEBSERVICE ksms562:9000 Server-Plugin-Alias: BMC.ARF.CMDBCONSOLE BMC.ARF.CMDBCONSOLE ksms562:9000 Server-Plugin-Alias: REMEDY.ARDBC.APPQUERY REMEDY.ARDBC.APPQUERY ksms562:9000 Server-Plugin-Alias: REMEDY.ARDBC.FLASHBOARD.FORM REMEDY.ARDB
AW: ARServer group and the ports that require load balancing
George, IMHO this is the relevant part of the documentation. Specifying plug-in server aliases (Step 4e) When you specify a common AR System server alias, AR System uses the alias name to connect to the plug-in server. When the AR System server is connecting to its local plug-in server, the connection will not be successful if the server name resolves to anything other than the local server. To route connections successfully in an environment where the server name resolves to a load balancer, you must define plug-in server aliases. The purpose of this step is to designate the local server as the host, not to change the name of the plug-in. To specify plug-in server aliases Add an alias definition for each loaded plug-in to the configuration file, using the plug-in name specified in the plug-in definition. Use the following format, substituting the alias name, plug-in name, host name, and optional port number: Server-Plugin-Alias: [:] The AR System server connects to the plug-in server using host at port when the AR System server communicates with the plug-in. Host must resolve to the local AR System server, and the port number is specified if the plug-in server is listening on a particular port. The following definition is a typical example: Server-Plugin-Alias: ARSYS.ARDBC.REPORT ARSYS.ARDBC.REPORT myserver:2020 For more information about Server-Plugin-Alias, see “ar.conf (ar.cfg)” on page 248. We have a similar setup and solved this problem in a different way. We added the following line to /etc/hosts 127.0.0.1 itsm.telekom.de itsm.telekom.de is configured as common server alias in ar.conf which is used to connect to the plugin server. HTH Kind Regards Conny Von: Action Request System discussion list(ARSList) [mailto:arsl...@arslist.org] Im Auftrag von George Jungle Gesendet: Dienstag, 7. April 2009 03:21 An: arslist@ARSLIST.ORG Betreff: ARServer group and the ports that require load balancing ** Guys, Wondering if you can help me out. We've got an issue with the Plug-in server running within a high available environment of Remedy (2 Midtiers, 2 AR Servers, 2 DB Servers) with each tier separated by a firewall. There is one load balancer in front of the midtier servers and another in front of the AR Servers. So, basically the AR Servers are running within a Server Group. When we try to open up a Remedy form, such as the "Server Information" form, we get an error saying : ARERR [8939] The AR System Plug-In server is not responding. Cannot connect to the system at this time. Contact your AR System Administrator for assistance. : RPC: Timed out Currently the only port that is load balanced is the TCP port that the users access the AR Servers by. Should we also need to load balance the Plug-In port? (9000)? All help is appreciated. Below are some emails I've dug up just from searching the web... Thanks. George Re: RPC: Unable to receive Tony Worthington Mon, 06 Aug 2007 11:27:16 -0700 You probably need to review the BMC documents about plugins behind a VIP or firewall without a portmapper. "Using a Hardware Load Balancer with BMC Remedy Action Request System 7.0" page 9, and others that I can't find at the moment. Here's an example of some lines from one of our ar.conf's (this one from a server group, but relevant since we don't use portmapper either) Plugin-Port: 9000 Register-With-Portmapper: F TCD-Specific-Port: 8639 Server-Plugin-Alias: ARSYS.ARDBC.CONFIGFILE ARSYS.ARDBC.CONFIGFILE ksms562:9000 Server-Plugin-Alias: ARSYS.ARDBC.LDAP ARSYS.ARDBC.LDAP ksms562:9000 Server-Plugin-Alias: ARSYS.ARDBC.PREVIEW ARSYS.ARDBC.PREVIEW ksms562:9000 Server-Plugin-Alias: ARSYS.AREA.LDAP ARSYS.AREA.LDAP ksms562:9000 Server-Plugin-Alias: ARSYS.ARF.WEBSERVICE ARSYS.ARF.WEBSERVICE ksms562:9000 Server-Plugin-Alias: BMC.ARF.CMDBCONSOLE BMC.ARF.CMDBCONSOLE ksms562:9000 Server-Plugin-Alias: REMEDY.ARDBC.APPQUERY REMEDY.ARDBC.APPQUERY ksms562:9000 Server-Plugin-Alias: REMEDY.ARDBC.FLASHBOARD.FORM REMEDY.ARDBC.FLASHBOARD.FORM ksms562:9000 Server-Plugin-Alias: REMEDY.ARDBC.SERVER.ADMINISTRATION REMEDY.ARDBC.SERVER.ADMINISTRATION ksms562:9000 Server-Plugin-Alias: REMEDY.ARF.CAI REMEDY.ARF.CAI ksms562:9000 Server-Plugin-Alias: REMEDY.ARF.CBDATA REMEDY.ARF.CBDATA ksms562:9000 Server-Plugin-Alias: REMEDY.ARF.OMFOBJIE REMEDY.ARF.OMFOBJIE ksms562:9000 Server-Plugin-Alias: REMEDY.ARF.SLASETUP REMEDY.ARF.SLASETUP ksms562:9000 You may not need the alias lines, but you certainly need the Plugin-Port. hth, tony -- Tony Worthington [EMAIL PROTECTED] 262-703-5911 Mike Chepaykin <[EMAIL PROTECTED]> Sent by: "Action Request System discussion list(ARSList)" 08/06/2007 10:12 AM Please respond to arslist@ARSLIST.ORG To arslist@ARSLIST.
ARServer group and the ports that require load balancing
Guys, Wondering if you can help me out. We've got an issue with the Plug-in server running within a high available environment of Remedy (2 Midtiers, 2 AR Servers, 2 DB Servers) with each tier separated by a firewall. There is one load balancer in front of the midtier servers and another in front of the AR Servers. So, basically the AR Servers are running within a Server Group. When we try to open up a Remedy form, such as the "Server Information" form, we get an error saying : ARERR [8939] The AR System Plug-In server is not responding. Cannot connect to the system at this time. Contact your AR System Administrator for assistance. : RPC: Timed out Currently the only port that is load balanced is the TCP port that the users access the AR Servers by. Should we also need to load balance the Plug-In port? (9000)? All help is appreciated. Below are some emails I've dug up just from searching the web... Thanks. George Re: RPC: Unable to receive Tony Worthington Mon, 06 Aug 2007 11:27:16 -0700 You probably need to review the BMC documents about plugins behind a VIP or firewall without a portmapper. "Using a Hardware Load Balancer with BMC Remedy Action Request System 7.0" page 9, and others that I can't find at the moment. Here's an example of some lines from one of our ar.conf's (this one from a server group, but relevant since we don't use portmapper either) Plugin-Port: 9000 Register-With-Portmapper: F TCD-Specific-Port: 8639 Server-Plugin-Alias: ARSYS.ARDBC.CONFIGFILE ARSYS.ARDBC.CONFIGFILE ksms562:9000 Server-Plugin-Alias: ARSYS.ARDBC.LDAP ARSYS.ARDBC.LDAP ksms562:9000 Server-Plugin-Alias: ARSYS.ARDBC.PREVIEW ARSYS.ARDBC.PREVIEW ksms562:9000 Server-Plugin-Alias: ARSYS.AREA.LDAP ARSYS.AREA.LDAP ksms562:9000 Server-Plugin-Alias: ARSYS.ARF.WEBSERVICE ARSYS.ARF.WEBSERVICE ksms562:9000 Server-Plugin-Alias: BMC.ARF.CMDBCONSOLE BMC.ARF.CMDBCONSOLE ksms562:9000 Server-Plugin-Alias: REMEDY.ARDBC.APPQUERY REMEDY.ARDBC.APPQUERY ksms562:9000 Server-Plugin-Alias: REMEDY.ARDBC.FLASHBOARD.FORM REMEDY.ARDBC.FLASHBOARD.FORM ksms562:9000 Server-Plugin-Alias: REMEDY.ARDBC.SERVER.ADMINISTRATION REMEDY.ARDBC.SERVER.ADMINISTRATION ksms562:9000 Server-Plugin-Alias: REMEDY.ARF.CAI REMEDY.ARF.CAI ksms562:9000 Server-Plugin-Alias: REMEDY.ARF.CBDATA REMEDY.ARF.CBDATA ksms562:9000 Server-Plugin-Alias: REMEDY.ARF.OMFOBJIE REMEDY.ARF.OMFOBJIE ksms562:9000 Server-Plugin-Alias: REMEDY.ARF.SLASETUP REMEDY.ARF.SLASETUP ksms562:9000 You may not need the alias lines, but you certainly need the Plugin-Port. hth, tony -- Tony Worthington [EMAIL PROTECTED] 262-703-5911 Mike Chepaykin <[EMAIL PROTECTED]> Sent by: "Action Request System discussion list(ARSList)" 08/06/2007 10:12 AM Please respond to arslist@ARSLIST.ORG To arslist@ARSLIST.ORG cc Subject RPC: Unable to receive Hello listers. ARERR [8760] Cannot establish a network connection to the AR System Plug-In server : bmcstage : RPC: Rpcbind failure - RPC: Unable to receive This is what I got in RUT when I move my ARS server behind the firewall, uncheck "Register with portmapper", assign static TCP port and shutdown remedy portmapper service. And restart my windows box of course - a shut in a head 8). This error appears only in Problem Management Console on window open. Other consoles - incident, asset, cmdb - appear without any errors. No errors in arerror, armonitor, arplugin logs and no errors on client side logs too. When I change environment back (no firewall restrictions at all) and move to portmapper - everything is fine again. Why (and which?) plugin try to register himself in portmapper? Or maybe the cause is something else? windows2003, mssql2000, ars7.0.1p2, ITSM 7.0.2 -- Mike Chepaykin ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: RMI Solutions ARSlist: "Where the Answers Are"
Re: MIDTIER - load balancing/clustering solutions
The mid-tier is not written for it. All the classes that make up the mid-tier do not support serialization. Axton The opinions, statements, and/or suggested courses of action expressed in this E-mail do not necessarily reflect those of BMC Software, Inc. My voluntary participation in this forum is not intended to convey a role as a spokesperson, liaison or public relations representative for BMC Software, Inc. On Tue, Dec 30, 2008 at 3:00 PM, Leihkauff, Kenneth < kenneth.g.leihka...@saic.com> wrote: > ** > > Hi, > > > > Has anyone had any experience using ARSystem Mid-tier in a clustered Tomcat > environment? > > Does anyone know if Mid-tier supports the inherent clustering that is built > into Tomcat? > > > > Thanks. > > > > Ken > > > __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" > html___ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: MIDTIER - load balancing/clustering solutions
We've been told by BMC that they do not support software load balancing. Steve Michadick Remedy Engineer MCNOSC -Original Message- From: Leihkauff, Kenneth [mailto:kenneth.g.leihka...@saic.com] Sent: Tuesday, December 30, 2008 4:01 PM Subject: MIDTIER - load balancing/clustering solutions ** Hi, Has anyone had any experience using ARSystem Mid-tier in a clustered Tomcat environment? Does anyone know if Mid-tier supports the inherent clustering that is built into Tomcat? Thanks. Ken __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" html___ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: MIDTIER - load balancing/clustering solutions
Ken, In our environment, we have each mid tier server running on its own instance of tomcat. We have Citrix Netscalers that do the loadbalancing in front of the mid tier servers. We then just configured the index.html on ROOT to redirect to arsys/home. Tauf Chowdhury | Forest Laboratories, Inc. Sr. Analyst Office: 631.858.7765 From: Action Request System discussion list(ARSList) [mailto:arsl...@arslist.org] On Behalf Of Leihkauff, Kenneth Sent: Tuesday, December 30, 2008 4:01 PM To: arslist@ARSLIST.ORG Subject: MIDTIER - load balancing/clustering solutions ** Hi, Has anyone had any experience using ARSystem Mid-tier in a clustered Tomcat environment? Does anyone know if Mid-tier supports the inherent clustering that is built into Tomcat? Thanks. Ken __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" html___ ** This e-mail and its attachments may contain Forest Laboratories, Inc. proprietary information that is privileged, confidential or subject to copyright belonging to Forest Laboratories, Inc. This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, or the employee or agent responsible for delivering this e-mail to the intended recipient, you are hereby notified that any dissemination, distribution, copying or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
MIDTIER - load balancing/clustering solutions
Hi, Has anyone had any experience using ARSystem Mid-tier in a clustered Tomcat environment? Does anyone know if Mid-tier supports the inherent clustering that is built into Tomcat? Thanks. Ken ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Very strange issue in assign trouble ticket with load balancing
Dear List , I got strange issue when I'm trying to assign trouble ticket without choosing the assignees ( Load Balancing ) , the assigne didnt appear in the system sometime its take 5 minutes sometimes , didnt appear at all ? Ar Server 7.1 Patch 4 Service Desk 7.3 Regards, Ramy ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Very strange issue in assign trouble ticket with load balancing
Dear List , I got strange issue when I'm trying to assign trouble ticket without choosing the assignees ( Load Balancing ) Ar Server 7.1 Patch 4 Service Desk 7.3 Regards, Ramy ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Very strange issue in assign trouble ticket with load balancing
Dear List , I got strange issue when I'm trying to assign trouble ticket without choosing the assignees ( Load Balancing ) Ar Server 7.1 Patch 4 Service Desk 7.3 Regards, Ramy ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Server Group - Load Balancing Questions
Yup. Would be cool to have a method to push users from one server to another in the group (with no interruption in service.) -- Tony Worthington Sr. Technical Analyst Kohl's Department Stores [EMAIL PROTECTED] 262-703-5911 Rick Cook <[EMAIL PROTECTED]> Sent by: "Action Request System discussion list(ARSList)" 05/29/2008 01:14 PM Please respond to arslist@ARSLIST.ORG To arslist@ARSLIST.ORG cc Subject Re: Server Group - Load Balancing Questions ** Something to know if you're new to Remedy LB is that it doesn't truly load-balance, defined as keeping the load balanced equally between all servers in the group. It's really load sharing. You might have 50 users over 3 servers, and depending on when they logged in, what they're doing, etc., you might have 13 on Server A, 30 on Server B, and 7 on Server C. That's probably not a real problem for most of us, but you should know to expect this going in. Rick On Thu, May 29, 2008 at 11:03 AM, Tony Worthington < [EMAIL PROTECTED]> wrote: ** Kelly Deaver did a great presentation at RUG on this topic and the presentation is available on the BMC/DN site. Load Balancing and High Availability for BMC Remedy Action Request System http://developer.bmc.com:8080/jiveProd/entry!default.jspa?categoryID=859&externalID=2841&fromSearchPage=true If I recall, you'll need to register and log in to view the attachment. There are also documents and whitepapers on the BMC support/documentation site. >Can I put a load balancer on top of a Remedy Server Group? Yes. >What is the recommended load balancing technology? There are quite a few solutions, including Alteon, F5, and many others. Use whatever your organization is familiar with. :-) >Is there any special Remedy licensing to utilize Server Group technology? Each server in the group needs an AR server license, but can share application and user licensing. You don't mention what version, but I would recommend 7.1 -- Tony Worthington Sr. Technical Analyst Kohl's Department Stores [EMAIL PROTECTED] 262-703-5911 "Wilson, Bruce B" <[EMAIL PROTECTED]> Sent by: "Action Request System discussion list(ARSList)" < arslist@ARSLIST.ORG> 05/29/2008 12:56 PM Please respond to arslist@ARSLIST.ORG To arslist@ARSLIST.ORG cc Subject Server Group - Load Balancing Questions ** I read about specifying the "Share" option when specifying the AR System database when installing AR System Server. I want to use the "Share" option to configure two Remedy servers to reside in a Remedy Server Group. Can I put a load balancer on top of a Remedy Server Group? What is the recommended load balancing technology? Is there anything that is not allowed? We are considering using "Windows Load Balancer". Is there any special Remedy licensing to utilize Server Group technology? Thanks, Bruce Wilson Datastream Consulting, LLC __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" html___ __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" html___ __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" html___ CONFIDENTIALITY NOTICE: This is a transmission from Kohl's Department Stores, Inc. and may contain information which is confidential and proprietary. If you are not the addressee, any disclosure, copying or distribution or use of the contents of this message is expressly prohibited. If you have received this transmission in error, please destroy it and notify us immediately at 262-703-7000. CAUTION: Internet and e-mail communications are Kohl's property and Kohl's reserves the right to retrieve and read any message created, sent and received. Kohl's reserves the right to monitor messages by authorized Kohl's Associates at any time without any further consent. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Server Group - Load Balancing Questions
Something to know if you're new to Remedy LB is that it doesn't truly load-balance, defined as keeping the load balanced equally between all servers in the group. It's really load sharing. You might have 50 users over 3 servers, and depending on when they logged in, what they're doing, etc., you might have 13 on Server A, 30 on Server B, and 7 on Server C. That's probably not a real problem for most of us, but you should know to expect this going in. Rick On Thu, May 29, 2008 at 11:03 AM, Tony Worthington < [EMAIL PROTECTED]> wrote: > ** > Kelly Deaver did a great presentation at RUG on this topic and the > presentation is available on the BMC/DN site. > > Load Balancing and High Availability for BMC Remedy Action Request System > > > > http://developer.bmc.com:8080/jiveProd/entry!default.jspa?categoryID=859&externalID=2841&fromSearchPage=true<http://developer.bmc.com:8080/jiveProd/entry%21default.jspa?categoryID=859&externalID=2841&fromSearchPage=true> > > If I recall, you'll need to register and log in to view the attachment. > There are also documents and whitepapers on the BMC support/documentation > site. > > >Can I put a load balancer on top of a Remedy Server Group? > Yes. > > >What is the recommended load balancing technology? > There are quite a few solutions, including Alteon, F5, and many others. > Use whatever your organization is familiar with. :-) > > >Is there any special Remedy licensing to utilize Server Group technology? > > Each server in the group needs an AR server license, but can share > application and user licensing. > > You don't mention what version, but I would recommend 7.1 > > -- > Tony Worthington > Sr. Technical Analyst > Kohl's Department Stores > [EMAIL PROTECTED] > 262-703-5911 > > > *"Wilson, Bruce B" <[EMAIL PROTECTED]>* > Sent by: "Action Request System discussion list(ARSList)" < > arslist@ARSLIST.ORG> > > 05/29/2008 12:56 PM > Please respond to > arslist@ARSLIST.ORG > > To > arslist@ARSLIST.ORG cc > Subject > Server Group - Load Balancing Questions > > > > > ** > I read about specifying the "Share" option when specifying the AR System > database when installing AR System Server. > I want to use the "Share" option to configure two Remedy servers to reside > in a Remedy Server Group. > > Can I put a load balancer on top of a Remedy Server Group? > What is the recommended load balancing technology? > Is there anything that is not allowed? We are considering using "Windows > Load Balancer". > > Is there any special Remedy licensing to utilize Server Group technology? > > Thanks, > Bruce Wilson > Datastream Consulting, LLC > > > __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" > html___ > > __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" > html___ > ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Server Group - Load Balancing Questions
Thanks Tony! I wish I had this slideshow earlier! ;) We have load balancing running on top of 2 ARS server sharing one database. We us BipIP for our load balancing. Tony Worthington <[EMAIL PROTECTED]> Sent by: "Action Request System discussion list(ARSList)" 05/29/2008 01:05 PM Please respond to arslist@ARSLIST.ORG To arslist@ARSLIST.ORG cc Subject Re: Server Group - Load Balancing Questions ** Kelly Deaver did a great presentation at RUG on this topic and the presentation is available on the BMC/DN site. Load Balancing and High Availability for BMC Remedy Action Request System http://developer.bmc.com:8080/jiveProd/entry!default.jspa?categoryID=859&externalID=2841&fromSearchPage=true If I recall, you'll need to register and log in to view the attachment. There are also documents and whitepapers on the BMC support/documentation site. >Can I put a load balancer on top of a Remedy Server Group? Yes. >What is the recommended load balancing technology? There are quite a few solutions, including Alteon, F5, and many others. Use whatever your organization is familiar with. :-) >Is there any special Remedy licensing to utilize Server Group technology? Each server in the group needs an AR server license, but can share application and user licensing. You don't mention what version, but I would recommend 7.1 -- Tony Worthington Sr. Technical Analyst Kohl's Department Stores [EMAIL PROTECTED] 262-703-5911 "Wilson, Bruce B" <[EMAIL PROTECTED]> Sent by: "Action Request System discussion list(ARSList)" 05/29/2008 12:56 PM Please respond to arslist@ARSLIST.ORG To arslist@ARSLIST.ORG cc Subject Server Group - Load Balancing Questions ** I read about specifying the ?Share? option when specifying the AR System database when installing AR System Server. I want to use the ?Share? option to configure two Remedy servers to reside in a Remedy Server Group. Can I put a load balancer on top of a Remedy Server Group? What is the recommended load balancing technology? Is there anything that is not allowed? We are considering using ?Windows Load Balancer?. Is there any special Remedy licensing to utilize Server Group technology? Thanks, Bruce Wilson Datastream Consulting, LLC __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" html___ __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" html___ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Server Group - Load Balancing Questions
answered inline _ From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Wilson, Bruce B Sent: Thursday, May 29, 2008 11:57 AM To: arslist@ARSLIST.ORG Subject: Server Group - Load Balancing Questions ** I read about specifying the "Share" option when specifying the AR System database when installing AR System Server. I want to use the "Share" option to configure two Remedy servers to reside in a Remedy Server Group. Can I put a load balancer on top of a Remedy Server Group? - Yes What is the recommended load balancing technology? - Only requirement that I'm familiar with is that it have the ability to do sticky bits Is there anything that is not allowed? - Not that I'm aware of We are considering using "Windows Load Balancer" Is there any special Remedy licensing to utilize Server Group technology? - Depends on your version...in 7.1 you don't need any special licensing...in previous versions you need to get a special key for your floating licenses. Thanks, Bruce Wilson Datastream Consulting, LLC __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" html___ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Server Group - Load Balancing Questions
Kelly Deaver did a great presentation at RUG on this topic and the presentation is available on the BMC/DN site. Load Balancing and High Availability for BMC Remedy Action Request System http://developer.bmc.com:8080/jiveProd/entry!default.jspa?categoryID=859&externalID=2841&fromSearchPage=true If I recall, you'll need to register and log in to view the attachment. There are also documents and whitepapers on the BMC support/documentation site. >Can I put a load balancer on top of a Remedy Server Group? Yes. >What is the recommended load balancing technology? There are quite a few solutions, including Alteon, F5, and many others. Use whatever your organization is familiar with. :-) >Is there any special Remedy licensing to utilize Server Group technology? Each server in the group needs an AR server license, but can share application and user licensing. You don't mention what version, but I would recommend 7.1 -- Tony Worthington Sr. Technical Analyst Kohl's Department Stores [EMAIL PROTECTED] 262-703-5911 "Wilson, Bruce B" <[EMAIL PROTECTED]> Sent by: "Action Request System discussion list(ARSList)" 05/29/2008 12:56 PM Please respond to arslist@ARSLIST.ORG To arslist@ARSLIST.ORG cc Subject Server Group - Load Balancing Questions ** I read about specifying the ?Share? option when specifying the AR System database when installing AR System Server. I want to use the ?Share? option to configure two Remedy servers to reside in a Remedy Server Group. Can I put a load balancer on top of a Remedy Server Group? What is the recommended load balancing technology? Is there anything that is not allowed? We are considering using ?Windows Load Balancer?. Is there any special Remedy licensing to utilize Server Group technology? Thanks, Bruce Wilson Datastream Consulting, LLC __Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" html___ CONFIDENTIALITY NOTICE: This is a transmission from Kohl's Department Stores, Inc. and may contain information which is confidential and proprietary. If you are not the addressee, any disclosure, copying or distribution or use of the contents of this message is expressly prohibited. If you have received this transmission in error, please destroy it and notify us immediately at 262-703-7000. CAUTION: Internet and e-mail communications are Kohl's property and Kohl's reserves the right to retrieve and read any message created, sent and received. Kohl's reserves the right to monitor messages by authorized Kohl's Associates at any time without any further consent. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Server Group - Load Balancing Questions
I read about specifying the "Share" option when specifying the AR System database when installing AR System Server. I want to use the "Share" option to configure two Remedy servers to reside in a Remedy Server Group. Can I put a load balancer on top of a Remedy Server Group? What is the recommended load balancing technology? Is there anything that is not allowed? We are considering using "Windows Load Balancer". Is there any special Remedy licensing to utilize Server Group technology? Thanks, Bruce Wilson Datastream Consulting, LLC ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Load Balancing
Thanks for all the responses everyone. I just grabbed that whitepaper and got a lot of other good information as well. The list makes me happy :) Michael A. McManus, SSgt, USAF Remedy Developer HQ 754 ELSG/DOMH From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Robert Page Sent: Thursday, December 13, 2007 3:13 AM To: arslist@ARSLIST.ORG Subject: Re: Load Balancing Michael I would seriously recommend reading the whitepaper on using a hardware load balancer with ar system, it has some useful pointers on load balancer and ar configurations and has some directions and instructions for server group settings. You should be able to get it form the support site. Regards Robert - Original Message From: McManus Michael A SSgt HQ 754 ELSG/DOMH <[EMAIL PROTECTED]> To: arslist@ARSLIST.ORG Sent: Wednesday, 12 December, 2007 8:25:28 PM Subject: Load Balancing ** Dear listers, We're in the process of a 7.0 upgrade and a new hardware configuration and we're trying to explore our options. I figured this is likely the best place to ask the pertinent questions. We just bought 3 new servers, and the plan was to set up a server group to help with load balancing. My question is what is the best way to handle something like this? We're totally new to this (we've always run on one production server with the database on the server) and our initial thought was 2 production servers with the user load split between the two and the third server as an external database. I was posed a question today and I'm unsure whether it can be done. Is it possible (and advisable if it is) to set up two production servers with a database on each server and have those databases effectively mirror each other in real time? I'm under the impression that a server grouping reduces stress on the server, but there is still a bottleneck at the database since all the servers in the group are using the same db. Anyone out there have experience with something like this that can point me in the right direction? Thanks much, Michael A. McManus, SSgt, USAF Remedy Developer HQ 754 ELSG/DOMH __20060125___This posting was submitted with HTML in it___ __20060125___This posting was submitted with HTML in it___ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Load Balancing
Michael I would seriously recommend reading the whitepaper on using a hardware load balancer with ar system, it has some useful pointers on load balancer and ar configurations and has some directions and instructions for server group settings. You should be able to get it form the support site. Regards Robert - Original Message From: McManus Michael A SSgt HQ 754 ELSG/DOMH <[EMAIL PROTECTED]> To: arslist@ARSLIST.ORG Sent: Wednesday, 12 December, 2007 8:25:28 PM Subject: Load Balancing ** Dear listers, We’re in the process of a 7.0 upgrade and a new hardware configuration and we’re trying to explore our options. I figured this is likely the best place to ask the pertinent questions. We just bought 3 new servers, and the plan was to set up a server group to help with load balancing. My question is what is the best way to handle something like this? We’re totally new to this (we’ve always run on one production server with the database on the server) and our initial thought was 2 production servers with the user load split between the two and the third server as an external database. I was posed a question today and I’m unsure whether it can be done. Is it possible (and advisable if it is) to set up two production servers with a database on each server and have those databases effectively mirror each other in real time? I’m under the impression that a server grouping reduces stress on the server, but there is still a bottleneck at the database since all the servers in the group are using the same db. Anyone out there have experience with something like this that can point me in the right direction? Thanks much, Michael A. McManus, SSgt, USAF Remedy Developer HQ 754 ELSG/DOMH __20060125___This posting was submitted with HTML in it___ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Load Balancing
And if you're using ITSM 7, increase your max thread count to about 2x-3x what you have been accustomed to using. It apparently does make a difference. Rick On 12/12/07, Hall Chad - chahal <[EMAIL PROTECTED]> wrote: > > ** > > You should go with 2 load balanced AR Servers in a server group with a > separate standalone database server. Generally AR Server is going to be your > bottleneck. A single database server should be able to easily handle the > load of multiple AR Servers. Just be sure to configure your AR Servers for > peak performance (fast/list/private threads, proper indexes, etc) > > > > *Chad Hall* > (501) 342-2650 > -- > > *From:* Action Request System discussion list(ARSList) [mailto: > [EMAIL PROTECTED] *On Behalf Of *McManus Michael A SSgt HQ 754 > ELSG/DOMH > *Sent:* Wednesday, December 12, 2007 2:25 PM > *To:* arslist@ARSLIST.ORG > *Subject:* Load Balancing > > > > Dear listers, > > We're in the process of a 7.0 upgrade and a new hardware > configuration and we're trying to explore our options. I figured this is > likely the best place to ask the pertinent questions. We just bought 3 new > servers, and the plan was to set up a server group to help with load > balancing. My question is what is the best way to handle something like > this? We're totally new to this (we've always run on one production server > with the database on the server) and our initial thought was 2 production > servers with the user load split between the two and the third server as an > external database. I was posed a question today and I'm unsure whether it > can be done. Is it possible (and advisable if it is) to set up two > production servers with a database on each server and have those databases > effectively mirror each other in real time? I'm under the impression that a > server grouping reduces stress on the server, but there is still a > bottleneck at the database since all the servers in the group are using the > same db. Anyone out there have experience with something like this that can > point me in the right direction? > > > > Thanks much, > > > > Michael A. McManus, SSgt, USAF > > Remedy Developer > > HQ 754 ELSG/DOMH > > *** > The information contained in this communication is confidential, is > intended only for the use of the recipient named above, and may be legally > privileged. > > If the reader of this message is not the intended recipient, you are > hereby notified that any dissemination, distribution or copying of this > communication is strictly prohibited. > > If you have received this communication in error, please resend this > communication to the sender and delete the original message or any copy > of it from your computer system. > > Thank You. > > > __20060125___This posting was submitted with HTML in > it___ __20060125___This posting was submitted with HTML > in it___ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Load Balancing
On server groups, all the application servers point to a central database. Currently, we have 2 production application servers that are load balanced through our F5 load balancer. Both servers are pointed to a central oracle database that runs on a different server. In your situation, you could keep the oracle database on the first server and just point the second server to that database or just move the database to a different server and have both application servers point to that database. A couple of things that I've noticed since we started the server group: - Check your load balancer to understand how it actually load balances. Ours for instance, balances based on IP's. Our problem resulting of this, is that all of our call center agents are NAT'd into one IP address while the rest of the company is NAT'd to a couple of others. Kind of defeats the purpose. - Make sure that the admin tool can modify workflow on both servers. With the load balancer, your not always going to get the same server (personal experience there). - Make sure that when you name your Remedy Application, that it is different accoss both servers. For example, if you have your first server name remedy and your second server is name remedy, you will not see both servers in the AR System Server Group Ranking form (personal experience there). - Make sure you set your AR System Server Group Rank form accordingly so that processes are not duplicated. Remember that both servers are pointing to the same database and hence the same workflow. This comes into view when running escalations. Currently, we have our first server doing all the e-mailing, escalations, and any other admin processes while the second server isn't assigned any tasks. - You may experience a license issue when you recieve your ServerGroup licenses from BMC. It's a bug that is hard to reproduce from their end apparently. We had to delete our .multilicense file and have the arsystem regenerate it. It's an annoying bugger. - Be aware that it took BMC several weeks to get our licenses and even then we had to resubmit several times because there was something wrong with them. When you do try to switch the servers over into ServerGroup mode make sure and keep the existing licenses in there so that you can roll back. We did this several times until we were sure all licenses were working correctly on both servers. Hopefully this helps you guys in your quest to have a server group. On Dec 12, 2007 1:25 PM, McManus Michael A SSgt HQ 754 ELSG/DOMH < [EMAIL PROTECTED]> wrote: > ** > > Dear listers, > > We're in the process of a 7.0 upgrade and a new hardware > configuration and we're trying to explore our options. I figured this is > likely the best place to ask the pertinent questions. We just bought 3 new > servers, and the plan was to set up a server group to help with load > balancing. My question is what is the best way to handle something like > this? We're totally new to this (we've always run on one production server > with the database on the server) and our initial thought was 2 production > servers with the user load split between the two and the third server as an > external database. I was posed a question today and I'm unsure whether it > can be done. Is it possible (and advisable if it is) to set up two > production servers with a database on each server and have those databases > effectively mirror each other in real time? I'm under the impression that a > server grouping reduces stress on the server, but there is still a > bottleneck at the database since all the servers in the group are using the > same db. Anyone out there have experience with something like this that can > point me in the right direction? > > > > Thanks much, > > > > Michael A. McManus, SSgt, USAF > > Remedy Developer > > HQ 754 ELSG/DOMH > __20060125___This posting was submitted with HTML in > it___ -- "A fool acts, regardless; knowing well that he is wrong. The ignoramus acts on only what he knows, but all that he knows. The ignoramus may be saved, but the fool knows that he is doomed." Robert Halstead ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Load Balancing
You should go with 2 load balanced AR Servers in a server group with a separate standalone database server. Generally AR Server is going to be your bottleneck. A single database server should be able to easily handle the load of multiple AR Servers. Just be sure to configure your AR Servers for peak performance (fast/list/private threads, proper indexes, etc) Chad Hall (501) 342-2650 From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of McManus Michael A SSgt HQ 754 ELSG/DOMH Sent: Wednesday, December 12, 2007 2:25 PM To: arslist@ARSLIST.ORG Subject: Load Balancing Dear listers, We're in the process of a 7.0 upgrade and a new hardware configuration and we're trying to explore our options. I figured this is likely the best place to ask the pertinent questions. We just bought 3 new servers, and the plan was to set up a server group to help with load balancing. My question is what is the best way to handle something like this? We're totally new to this (we've always run on one production server with the database on the server) and our initial thought was 2 production servers with the user load split between the two and the third server as an external database. I was posed a question today and I'm unsure whether it can be done. Is it possible (and advisable if it is) to set up two production servers with a database on each server and have those databases effectively mirror each other in real time? I'm under the impression that a server grouping reduces stress on the server, but there is still a bottleneck at the database since all the servers in the group are using the same db. Anyone out there have experience with something like this that can point me in the right direction? Thanks much, Michael A. McManus, SSgt, USAF Remedy Developer HQ 754 ELSG/DOMH __20060125___This posting was submitted with HTML in it___ *** The information contained in this communication is confidential, is intended only for the use of the recipient named above, and may be legally privileged. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please resend this communication to the sender and delete the original message or any copy of it from your computer system. Thank You. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Load Balancing
Dear listers, We're in the process of a 7.0 upgrade and a new hardware configuration and we're trying to explore our options. I figured this is likely the best place to ask the pertinent questions. We just bought 3 new servers, and the plan was to set up a server group to help with load balancing. My question is what is the best way to handle something like this? We're totally new to this (we've always run on one production server with the database on the server) and our initial thought was 2 production servers with the user load split between the two and the third server as an external database. I was posed a question today and I'm unsure whether it can be done. Is it possible (and advisable if it is) to set up two production servers with a database on each server and have those databases effectively mirror each other in real time? I'm under the impression that a server grouping reduces stress on the server, but there is still a bottleneck at the database since all the servers in the group are using the same db. Anyone out there have experience with something like this that can point me in the right direction? Thanks much, Michael A. McManus, SSgt, USAF Remedy Developer HQ 754 ELSG/DOMH ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
another load balancing question: Linux/Tomcat/Apache with Netscalers
All the diagrams I've ever seen only show one load balancer. Sure there is one, but there are two ports mapped. We have two load balancers in front of our backend web servers that run mid-tier. In the past we have been running CSM load balancers and we were told by Remedy Support to force all transactions to https so that the session ip stickiness would be on the same backend server after login. Apache / Tomcat 5.5 Linux RHEL-3 MidTier 6.3 We are trying to move to new Netscaler LB but there are problems. Port:80 --> backend port 80 on all web servers (Apache with rewrites to force all connections to https) Port:443 --> backen port 8443 running tomcat and midtier -- -Geoff Endresen Amazon.com ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Mid-Tier in Load Balancing
None really. Just make sure the sticky bit is enabled and that you use a least connections method of balancing. Also make sure the session timeout isn't smaller than your Mid Tier timeout. There are some occasional issues with the AR Server not allowing a user to login because they're already logged in from another machine. This seems to happen if a user connects through Mid Tier and gets routed to server A, then closes the browser without logging off (or they can even try to logoff but Mid Tier doesn't actually log them off). If they then try it again within 15 minutes and the load balancer (seeing a brand new connection) routes them to server B, then AR Server will prevent the login because it thinks they're still logged in from server A. We don't have too many issues with this, but then most of our users are on Remedy User. Sometimes when the issue happens a lot we reboot the web servers and it seems to correct itself. I'm hoping this is handled better in 7.1. Chad Hall (501) 342-2650 From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of LJ LongWing (Head) Sent: Monday, November 05, 2007 1:30 PM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing What sort of issues did you experience? From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Howard Richter Sent: Monday, November 05, 2007 11:56 AM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing ** Chad, We worked for 6 months to get the F5 to work with 7.0 mid-tier and gave up since our load was not to large.We are using IIS All I can say is good luck. Howard On 11/5/07, Hall Chad - chahal <[EMAIL PROTECTED]> wrote: We've used an F5 BigIP load balancer for two IIS servers for over 4 years without problems. They are currently running Mid Tier 6.3 patch 20 on IIS/ServletExec and are about to be rebuilt for Apache/Tomcat. But all the load balancing is done through the BigIP, not through IIS. We have a 60 minute timeout with session affinity (sticky bit) enabled and balancing based on least connections. Chad Hall (501) 342-2650 -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG ] On Behalf Of LJ LongWing (Head) Sent: Monday, November 05, 2007 10:40 AM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing What load balancer are you using? What are your session timeouts on session affinity? -Original Message- From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Geoffrey Endresen Sent: Saturday, November 03, 2007 11:42 PM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing We use apache. But we were told by support to route all traffic from http to https using rewrite commands in apache. Then, the ip stickiness on the https load balancer works. I don't like the solution. But it's the only one we have working for AR System 6.3 mid-tier Thanks, -Geoff On Nov 2, 2007 10:43 AM, LJ LongWing (Head) <[EMAIL PROTECTED]> wrote: > Mid-Tier 7.1 > Remedy 6.3 P21 > Tomcat (version?) > > Ok...Here is the situation. We are migrating from a App/Web on same > box to a Load Balanced web server with app on separate server (3 boxes in total). > They are using IIS Load Balancing with session affinity. When both > boxes are enable in the load balancer people are getting 9201 errors. > we have played with the load balancing config and verified that each > box individually is functioning properly without error. The only > thing that the Mid-Tier doc says about Load Balancing is to ensure > that you have sticky bits turned on. We are using SSL and the IIS > team is saying they can only track SSL sessions but don't know how to > associate that with a client session. We are using OOTB Tomcat that > came with the 7.1 Mid-Tier. Any suggestions of how to troubleshoot this issue? > > __ > _ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > ARSlist:"Where the Answers Are" > -- -Geoff Endresen Amazon.com ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" *** The information contained in this communication is confidential, is intended only for the use of the recipient n
Re: Mid-Tier in Load Balancing
In our case you had the f5 and then pointing to boxa and boxb. If the F5 was in use you would connect to boxa, but you could not log in. This was the same if for some reason you went to boxb. But if only boxa (or boxb) were on line everything worked. Our network team tried everyting and BMC was of no help. I could try and find some of the e-mails, but we gave up about 4 months ago. Howard On 11/5/07, LJ LongWing (Head) <[EMAIL PROTECTED]> wrote: > > ** What sort of issues did you experience? > > -- > *From:* Action Request System discussion list(ARSList) [mailto: > [EMAIL PROTECTED] *On Behalf Of *Howard Richter > *Sent:* Monday, November 05, 2007 11:56 AM > *To:* arslist@ARSLIST.ORG > *Subject:* Re: Mid-Tier in Load Balancing > > > ** Chad, > > We worked for 6 months to get the F5 to work with 7.0 mid-tier and gave up > since our load was not to large.We are using IIS > > All I can say is good luck. > > Howard > > > On 11/5/07, Hall Chad - chahal <[EMAIL PROTECTED]> wrote: > > > > We've used an F5 BigIP load balancer for two IIS servers for over 4 > > years without problems. They are currently running Mid Tier 6.3 patch 20 > > on IIS/ServletExec and are about to be rebuilt for Apache/Tomcat. But > > all the load balancing is done through the BigIP, not through IIS. We > > have a 60 minute timeout with session affinity (sticky bit) enabled and > > balancing based on least connections. > > > > Chad Hall > > (501) 342-2650 > > > > -Original Message- > > From: Action Request System discussion list(ARSList) > > [mailto:arslist@ARSLIST.ORG ] On Behalf Of LJ LongWing (Head) > > Sent: Monday, November 05, 2007 10:40 AM > > To: arslist@ARSLIST.ORG > > Subject: Re: Mid-Tier in Load Balancing > > > > What load balancer are you using? What are your session timeouts on > > session > > affinity? > > > > -Original Message- > > From: Action Request System discussion list(ARSList) > > [mailto:[EMAIL PROTECTED] On Behalf Of Geoffrey Endresen > > Sent: Saturday, November 03, 2007 11:42 PM > > To: arslist@ARSLIST.ORG > > Subject: Re: Mid-Tier in Load Balancing > > > > We use apache. But we were told by support to route all traffic from > > http to > > https using rewrite commands in apache. Then, the ip stickiness on the > > https > > load balancer works. I don't like the solution. But it's the only one we > > have working for AR System 6.3 mid-tier > > > > Thanks, > > -Geoff > > > > On Nov 2, 2007 10:43 AM, LJ LongWing (Head) <[EMAIL PROTECTED]> > > wrote: > > > Mid-Tier 7.1 > > > Remedy 6.3 P21 > > > Tomcat (version?) > > > > > > Ok...Here is the situation. We are migrating from a App/Web on same > > > box to a Load Balanced web server with app on separate server (3 boxes > > in > > total). > > > They are using IIS Load Balancing with session affinity. When both > > > boxes are enable in the load balancer people are getting 9201 errors. > > > > > we have played with the load balancing config and verified that each > > > box individually is functioning properly without error. The only > > > thing that the Mid-Tier doc says about Load Balancing is to ensure > > > that you have sticky bits turned on. We are using SSL and the IIS > > > team is saying they can only track SSL sessions but don't know how to > > > associate that with a client session. We are using OOTB Tomcat that > > > came with the 7.1 Mid-Tier. Any suggestions of how to troubleshoot > > this > > issue? > > > > > > __ > > > > > _ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > > > ARSlist:"Where the Answers Are" > > > > > > > > > > > -- > > -Geoff Endresen > > Amazon.com <http://amazon.com/> > > > > > > > > ___ > > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum > > Sponsor: > > www.rmsportal.com ARSlist: "Where the Answers Are" > > > > > > ___ > > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > > Platinum Sponsor: www.rmsp
Re: Mid-Tier in Load Balancing
We've been testing 7.0.1 with Apache/Tomcat for the last 3 months without any load balancing issues. We're moving it to production very soon. Chad Hall (501) 342-2650 From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Howard Richter Sent: Monday, November 05, 2007 12:56 PM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing ** Chad, We worked for 6 months to get the F5 to work with 7.0 mid-tier and gave up since our load was not to large.We are using IIS All I can say is good luck. Howard On 11/5/07, Hall Chad - chahal <[EMAIL PROTECTED]> wrote: We've used an F5 BigIP load balancer for two IIS servers for over 4 years without problems. They are currently running Mid Tier 6.3 patch 20 on IIS/ServletExec and are about to be rebuilt for Apache/Tomcat. But all the load balancing is done through the BigIP, not through IIS. We have a 60 minute timeout with session affinity (sticky bit) enabled and balancing based on least connections. Chad Hall (501) 342-2650 -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG ] On Behalf Of LJ LongWing (Head) Sent: Monday, November 05, 2007 10:40 AM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing What load balancer are you using? What are your session timeouts on session affinity? -Original Message- From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Geoffrey Endresen Sent: Saturday, November 03, 2007 11:42 PM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing We use apache. But we were told by support to route all traffic from http to https using rewrite commands in apache. Then, the ip stickiness on the https load balancer works. I don't like the solution. But it's the only one we have working for AR System 6.3 mid-tier Thanks, -Geoff On Nov 2, 2007 10:43 AM, LJ LongWing (Head) <[EMAIL PROTECTED]> wrote: > Mid-Tier 7.1 > Remedy 6.3 P21 > Tomcat (version?) > > Ok...Here is the situation. We are migrating from a App/Web on same > box to a Load Balanced web server with app on separate server (3 boxes in total). > They are using IIS Load Balancing with session affinity. When both > boxes are enable in the load balancer people are getting 9201 errors. > we have played with the load balancing config and verified that each > box individually is functioning properly without error. The only > thing that the Mid-Tier doc says about Load Balancing is to ensure > that you have sticky bits turned on. We are using SSL and the IIS > team is saying they can only track SSL sessions but don't know how to > associate that with a client session. We are using OOTB Tomcat that > came with the 7.1 Mid-Tier. Any suggestions of how to troubleshoot this issue? > > __ > _ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > ARSlist:"Where the Answers Are" > -- -Geoff Endresen Amazon.com ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" *** The information contained in this communication is confidential, is intended only for the use of the recipient named above, and may be legally privileged. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please resend this communication to the sender and delete the original message or any copy of it from your computer system. Thank You. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" -- Howard Richter ITIL Foundation Certified Red Hat Certified Technician CompTIA Linux+ Certified [EMAIL PROTECTED] Resume = http://hotjobs.yahoo.com/resumes/hrichter_1/resumeofhoward __20060125___This posting was submitted with HTML in it___ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Mid-Tier in Load Balancing
What sort of issues did you experience? _ From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Howard Richter Sent: Monday, November 05, 2007 11:56 AM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing ** Chad, We worked for 6 months to get the F5 to work with 7.0 mid-tier and gave up since our load was not to large.We are using IIS All I can say is good luck. Howard On 11/5/07, Hall Chad - chahal <[EMAIL PROTECTED]> wrote: We've used an F5 BigIP load balancer for two IIS servers for over 4 years without problems. They are currently running Mid Tier 6.3 patch 20 on IIS/ServletExec and are about to be rebuilt for Apache/Tomcat. But all the load balancing is done through the BigIP, not through IIS. We have a 60 minute timeout with session affinity (sticky bit) enabled and balancing based on least connections. Chad Hall (501) 342-2650 -Original Message- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG <mailto:arslist@ARSLIST.ORG> ] On Behalf Of LJ LongWing (Head) Sent: Monday, November 05, 2007 10:40 AM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing What load balancer are you using? What are your session timeouts on session affinity? -Original Message- From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Geoffrey Endresen Sent: Saturday, November 03, 2007 11:42 PM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing We use apache. But we were told by support to route all traffic from http to https using rewrite commands in apache. Then, the ip stickiness on the https load balancer works. I don't like the solution. But it's the only one we have working for AR System 6.3 mid-tier Thanks, -Geoff On Nov 2, 2007 10:43 AM, LJ LongWing (Head) <[EMAIL PROTECTED]> wrote: > Mid-Tier 7.1 > Remedy 6.3 P21 > Tomcat (version?) > > Ok...Here is the situation. We are migrating from a App/Web on same > box to a Load Balanced web server with app on separate server (3 boxes in total). > They are using IIS Load Balancing with session affinity. When both > boxes are enable in the load balancer people are getting 9201 errors. > we have played with the load balancing config and verified that each > box individually is functioning properly without error. The only > thing that the Mid-Tier doc says about Load Balancing is to ensure > that you have sticky bits turned on. We are using SSL and the IIS > team is saying they can only track SSL sessions but don't know how to > associate that with a client session. We are using OOTB Tomcat that > came with the 7.1 Mid-Tier. Any suggestions of how to troubleshoot this issue? > > __ > _ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > ARSlist:"Where the Answers Are" > -- -Geoff Endresen Amazon.com ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" *** The information contained in this communication is confidential, is intended only for the use of the recipient named above, and may be legally privileged. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please resend this communication to the sender and delete the original message or any copy of it from your computer system. Thank You. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" -- Howard Richter ITIL Foundation Certified Red Hat Certified Technician CompTIA Linux+ Certified [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> Resume = http://hotjobs.yahoo.com/resumes/hrichter_1/resumeofhoward __20060125___This posting was submitted with HTML in it___ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Mid-Tier in Load Balancing
Chad, We worked for 6 months to get the F5 to work with 7.0 mid-tier and gave up since our load was not to large.We are using IIS All I can say is good luck. Howard On 11/5/07, Hall Chad - chahal <[EMAIL PROTECTED]> wrote: > > We've used an F5 BigIP load balancer for two IIS servers for over 4 > years without problems. They are currently running Mid Tier 6.3 patch 20 > on IIS/ServletExec and are about to be rebuilt for Apache/Tomcat. But > all the load balancing is done through the BigIP, not through IIS. We > have a 60 minute timeout with session affinity (sticky bit) enabled and > balancing based on least connections. > > Chad Hall > (501) 342-2650 > > -Original Message- > From: Action Request System discussion list(ARSList) > [mailto:[EMAIL PROTECTED] On Behalf Of LJ LongWing (Head) > Sent: Monday, November 05, 2007 10:40 AM > To: arslist@ARSLIST.ORG > Subject: Re: Mid-Tier in Load Balancing > > What load balancer are you using? What are your session timeouts on > session > affinity? > > -Original Message- > From: Action Request System discussion list(ARSList) > [mailto:[EMAIL PROTECTED] On Behalf Of Geoffrey Endresen > Sent: Saturday, November 03, 2007 11:42 PM > To: arslist@ARSLIST.ORG > Subject: Re: Mid-Tier in Load Balancing > > We use apache. But we were told by support to route all traffic from > http to > https using rewrite commands in apache. Then, the ip stickiness on the > https > load balancer works. I don't like the solution. But it's the only one we > have working for AR System 6.3 mid-tier > > Thanks, > -Geoff > > On Nov 2, 2007 10:43 AM, LJ LongWing (Head) <[EMAIL PROTECTED]> > wrote: > > Mid-Tier 7.1 > > Remedy 6.3 P21 > > Tomcat (version?) > > > > Ok...Here is the situation. We are migrating from a App/Web on same > > box to a Load Balanced web server with app on separate server (3 boxes > in > total). > > They are using IIS Load Balancing with session affinity. When both > > boxes are enable in the load balancer people are getting 9201 errors. > > > we have played with the load balancing config and verified that each > > box individually is functioning properly without error. The only > > thing that the Mid-Tier doc says about Load Balancing is to ensure > > that you have sticky bits turned on. We are using SSL and the IIS > > team is saying they can only track SSL sessions but don't know how to > > associate that with a client session. We are using OOTB Tomcat that > > came with the 7.1 Mid-Tier. Any suggestions of how to troubleshoot > this > issue? > > > > __ > > _ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > > ARSlist:"Where the Answers Are" > > > > > > -- > -Geoff Endresen > Amazon.com > > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum > Sponsor: > www.rmsportal.com ARSlist: "Where the Answers Are" > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" > > *** > The information contained in this communication is confidential, is > intended only for the use of the recipient named above, and may be legally > privileged. > > If the reader of this message is not the intended recipient, you are > hereby notified that any dissemination, distribution or copying of this > communication is strictly prohibited. > > If you have received this communication in error, please resend this > communication to the sender and delete the original message or any copy > of it from your computer system. > > Thank You. > > > > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" > -- Howard Richter ITIL Foundation Certified Red Hat Certified Technician CompTIA Linux+ Certified [EMAIL PROTECTED] Resume = http://hotjobs.yahoo.com/resumes/hrichter_1/resumeofhoward ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Mid-Tier in Load Balancing
We've used an F5 BigIP load balancer for two IIS servers for over 4 years without problems. They are currently running Mid Tier 6.3 patch 20 on IIS/ServletExec and are about to be rebuilt for Apache/Tomcat. But all the load balancing is done through the BigIP, not through IIS. We have a 60 minute timeout with session affinity (sticky bit) enabled and balancing based on least connections. Chad Hall (501) 342-2650 -Original Message- From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of LJ LongWing (Head) Sent: Monday, November 05, 2007 10:40 AM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing What load balancer are you using? What are your session timeouts on session affinity? -Original Message- From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Geoffrey Endresen Sent: Saturday, November 03, 2007 11:42 PM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing We use apache. But we were told by support to route all traffic from http to https using rewrite commands in apache. Then, the ip stickiness on the https load balancer works. I don't like the solution. But it's the only one we have working for AR System 6.3 mid-tier Thanks, -Geoff On Nov 2, 2007 10:43 AM, LJ LongWing (Head) <[EMAIL PROTECTED]> wrote: > Mid-Tier 7.1 > Remedy 6.3 P21 > Tomcat (version?) > > Ok...Here is the situation. We are migrating from a App/Web on same > box to a Load Balanced web server with app on separate server (3 boxes in total). > They are using IIS Load Balancing with session affinity. When both > boxes are enable in the load balancer people are getting 9201 errors. > we have played with the load balancing config and verified that each > box individually is functioning properly without error. The only > thing that the Mid-Tier doc says about Load Balancing is to ensure > that you have sticky bits turned on. We are using SSL and the IIS > team is saying they can only track SSL sessions but don't know how to > associate that with a client session. We are using OOTB Tomcat that > came with the 7.1 Mid-Tier. Any suggestions of how to troubleshoot this issue? > > __ > _ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > ARSlist:"Where the Answers Are" > -- -Geoff Endresen Amazon.com ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" *** The information contained in this communication is confidential, is intended only for the use of the recipient named above, and may be legally privileged. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please resend this communication to the sender and delete the original message or any copy of it from your computer system. Thank You. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Mid-Tier in Load Balancing
What load balancer are you using? What are your session timeouts on session affinity? -Original Message- From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Geoffrey Endresen Sent: Saturday, November 03, 2007 11:42 PM To: arslist@ARSLIST.ORG Subject: Re: Mid-Tier in Load Balancing We use apache. But we were told by support to route all traffic from http to https using rewrite commands in apache. Then, the ip stickiness on the https load balancer works. I don't like the solution. But it's the only one we have working for AR System 6.3 mid-tier Thanks, -Geoff On Nov 2, 2007 10:43 AM, LJ LongWing (Head) <[EMAIL PROTECTED]> wrote: > Mid-Tier 7.1 > Remedy 6.3 P21 > Tomcat (version?) > > Ok...Here is the situation. We are migrating from a App/Web on same > box to a Load Balanced web server with app on separate server (3 boxes in total). > They are using IIS Load Balancing with session affinity. When both > boxes are enable in the load balancer people are getting 9201 errors. > we have played with the load balancing config and verified that each > box individually is functioning properly without error. The only > thing that the Mid-Tier doc says about Load Balancing is to ensure > that you have sticky bits turned on. We are using SSL and the IIS > team is saying they can only track SSL sessions but don't know how to > associate that with a client session. We are using OOTB Tomcat that > came with the 7.1 Mid-Tier. Any suggestions of how to troubleshoot this issue? > > __ > _ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > ARSlist:"Where the Answers Are" > -- -Geoff Endresen Amazon.com ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Re: Mid-Tier in Load Balancing
We use apache. But we were told by support to route all traffic from http to https using rewrite commands in apache. Then, the ip stickiness on the https load balancer works. I don't like the solution. But it's the only one we have working for AR System 6.3 mid-tier Thanks, -Geoff On Nov 2, 2007 10:43 AM, LJ LongWing (Head) <[EMAIL PROTECTED]> wrote: > Mid-Tier 7.1 > Remedy 6.3 P21 > Tomcat (version?) > > Ok...Here is the situation. We are migrating from a App/Web on same box to > a Load Balanced web server with app on separate server (3 boxes in total). > They are using IIS Load Balancing with session affinity. When both boxes > are enable in the load balancer people are getting 9201 errors. we have > played with the load balancing config and verified that each box > individually is functioning properly without error. The only thing that the > Mid-Tier doc says about Load Balancing is to ensure that you have sticky > bits turned on. We are using SSL and the IIS team is saying they can only > track SSL sessions but don't know how to associate that with a client > session. We are using OOTB Tomcat that came with the 7.1 Mid-Tier. Any > suggestions of how to troubleshoot this issue? > > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:"Where the > Answers Are" > -- -Geoff Endresen Amazon.com ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
Mid-Tier in Load Balancing
Mid-Tier 7.1 Remedy 6.3 P21 Tomcat (version?) Ok...Here is the situation. We are migrating from a App/Web on same box to a Load Balanced web server with app on separate server (3 boxes in total). They are using IIS Load Balancing with session affinity. When both boxes are enable in the load balancer people are getting 9201 errors. we have played with the load balancing config and verified that each box individually is functioning properly without error. The only thing that the Mid-Tier doc says about Load Balancing is to ensure that you have sticky bits turned on. We are using SSL and the IIS team is saying they can only track SSL sessions but don't know how to associate that with a client session. We are using OOTB Tomcat that came with the 7.1 Mid-Tier. Any suggestions of how to troubleshoot this issue? ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:"Where the Answers Are"
Re: Server Clustering and Load Balancing
We've used F5 Big IPs to load balance two AR Servers and 2 Mid Tier servers since 2003. It works great. Its configured to monitor a given port every 15 seconds to make sure the application server is alive (port 9030 for AR Server, port 443 for Mid Tier). If at any point a server doesn't respond it redirects user traffic to the other server in the group. Works very nice. Chad Hall (501) 342-2650 -Original Message- From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Robert Halstead Sent: Friday, July 27, 2007 4:31 PM To: arslist@ARSLIST.ORG Subject: Re: Server Clustering and Load Balancing We just recently got a server group up and running in our environment. We're using Big IP F5 to load balance between the two servers. So far we have had no problems load balencing between them except for the hardware itself. The F5 doesn't do fail over very well unless the whole server box goes down. On 7/27/07, Jase Brandon <[EMAIL PROTECTED]> wrote: > ** > > Hello All, > Due to growing performance issues, We are exploring the possibility of > moving our existing server to a clustered environment and utilizing load > balancing technologies. > I wondered if anyone has any best case scenarios/success stories or any > information at all regarding this topic. > > Thanks in Advance, > > > > Jase Brandon > > Remedy Administration/Development > > Customer Support Systems Group > > Desk - (615) - 320-4494 > > Cell - (334) - 318-5426 > > [EMAIL PROTECTED] > DaVita Inc. __20060125___This posting > was submitted with HTML in it___ -- "A fool acts, regardless; knowing well that he is wrong. The ignoramus acts on only what he knows, but all that he knows. The ignoramus may be saved, but the fool knows that he is doomed." Robert Halstead ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:"Where the Answers Are" * The information contained in this communication is confidential, is intended only for the use of the recipient named above, and may be legally privileged. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please resend this communication to the sender and delete the original message or any copy of it from your computer system. Thank you. * ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:"Where the Answers Are"
Re: Server Clustering and Load Balancing
We just recently got a server group up and running in our environment. We're using Big IP F5 to load balance between the two servers. So far we have had no problems load balencing between them except for the hardware itself. The F5 doesn't do fail over very well unless the whole server box goes down. On 7/27/07, Jase Brandon <[EMAIL PROTECTED]> wrote: > ** > > Hello All, > Due to growing performance issues, We are exploring the possibility of > moving our existing server to a clustered environment and utilizing load > balancing technologies. > I wondered if anyone has any best case scenarios/success stories or any > information at all regarding this topic. > > Thanks in Advance, > > > > Jase Brandon > > Remedy Administration/Development > > Customer Support Systems Group > > Desk - (615) - 320-4494 > > Cell - (334) - 318-5426 > > [EMAIL PROTECTED] > DaVita Inc. __20060125___This posting > was submitted with HTML in it___ -- "A fool acts, regardless; knowing well that he is wrong. The ignoramus acts on only what he knows, but all that he knows. The ignoramus may be saved, but the fool knows that he is doomed." Robert Halstead ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:"Where the Answers Are"
Re: Server Clustering and Load Balancing (U)
UNCLASSIFIED Our application and database Windows 2003 servers have been running on VERITAS clustered servers since Feb 2005 Sandra Hennigan Enterprise Remedy Administrator Office # 703-602-2525 x251 CACI - Ever Vigilant Apparently, there is nothing that cannot happen today. Mark Twain -Original Message- From: Action Request System discussion list(ARSList) [mailto:[EMAIL PROTECTED] On Behalf Of Jase Brandon Sent: Friday, July 27, 2007 1:30 PM To: arslist@ARSLIST.ORG Subject: Server Clustering and Load Balancing ** Hello All, Due to growing performance issues, We are exploring the possibility of moving our existing server to a clustered environment and utilizing load balancing technologies. I wondered if anyone has any best case scenarios/success stories or any information at all regarding this topic. Thanks in Advance, Jase Brandon Remedy Administration/Development Customer Support Systems Group Desk - (615) - 320-4494 Cell - (334) - 318-5426 [EMAIL PROTECTED] DaVita Inc. __20060125___This posting was submitted with HTML in it___ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:"Where the Answers Are"
Server Clustering and Load Balancing
Hello All, Due to growing performance issues, We are exploring the possibility of moving our existing server to a clustered environment and utilizing load balancing technologies. I wondered if anyone has any best case scenarios/success stories or any information at all regarding this topic. Thanks in Advance, Jase Brandon Remedy Administration/Development Customer Support Systems Group Desk - (615) - 320-4494 Cell - (334) - 318-5426 [EMAIL PROTECTED] DaVita Inc. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:"Where the Answers Are"