RE: Tomcat Alone or tomcat+IIS/Apache
> From: Gregg D Bolinger [mailto:[EMAIL PROTECTED] > Subject: Re: Tomcat Alone or tomcat+IIS/Apache > > Does Tomcat support CGI bins > utalizing non-java technology? As usual, RTFM: http://jakarta.apache.org/tomcat/tomcat-5.5-doc/cgi-howto.html - Chuck THIS COMMUNICATION MAY CONTAIN CONFIDENTIAL AND/OR OTHERWISE PROPRIETARY MATERIAL and is thus for use only by the intended recipient. If you received this in error, please contact the sender and delete the e-mail and its attachments from all computers. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Tomcat Alone or tomcat+IIS/Apache
We required IIS for a CGI ecommerce solution that was required. Plug and Pay I think is the company we bought it from. I came in late on the project and haven't had a lot of time to mess with it. Does Tomcat support CGI bins utalizing non-java technology? Gregg On 9/30/05, Peddireddy Srikanth <[EMAIL PROTECTED]> wrote: > > thanks for all the replies > > > On 9/30/05, Caldarale, Charles R <[EMAIL PROTECTED]> wrote: > > > From: Peddireddy Srikanth [mailto:[EMAIL PROTECTED] > > > Subject: Re: Tomcat Alone or tomcat+IIS/Apache > > > > > > And they argue that as Tomcat it self runs inside a JVM, which inturn > > > is a single process all the threads etc wil be simulted ones (and not > > > the native threads) and hence it will not scale up well under high > > > loads. > > > > More urban myth. As another respondent pointed out, all modern JVMs > > (i.e., from JRE 1.2 on) use native threads and the underlying OS for > > thread dispatching. Coupled with thread-local object allocation > > (available since JRE 1.3), scaling of Tomcat itself is not a problem; it > > runs happily on our 32-CPU servers as long as the applications > > themselves have no inherent bottlenecks. You do want to adjust the heap > > parameters for any serious work, especially on a Windows platform, where > > the default maximum borders on the miniscule. > > > > - Chuck > > > > > > THIS COMMUNICATION MAY CONTAIN CONFIDENTIAL AND/OR OTHERWISE PROPRIETARY > > MATERIAL and is thus for use only by the intended recipient. If you > > received this in error, please contact the sender and delete the e-mail > > and its attachments from all computers. > > > > - > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > - > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > >
Re: Tomcat Alone or tomcat+IIS/Apache
thanks for all the replies On 9/30/05, Caldarale, Charles R <[EMAIL PROTECTED]> wrote: > > From: Peddireddy Srikanth [mailto:[EMAIL PROTECTED] > > Subject: Re: Tomcat Alone or tomcat+IIS/Apache > > > > And they argue that as Tomcat it self runs inside a JVM, which inturn > > is a single process all the threads etc wil be simulted ones (and not > > the native threads) and hence it will not scale up well under high > > loads. > > More urban myth. As another respondent pointed out, all modern JVMs > (i.e., from JRE 1.2 on) use native threads and the underlying OS for > thread dispatching. Coupled with thread-local object allocation > (available since JRE 1.3), scaling of Tomcat itself is not a problem; it > runs happily on our 32-CPU servers as long as the applications > themselves have no inherent bottlenecks. You do want to adjust the heap > parameters for any serious work, especially on a Windows platform, where > the default maximum borders on the miniscule. > > - Chuck > > > THIS COMMUNICATION MAY CONTAIN CONFIDENTIAL AND/OR OTHERWISE PROPRIETARY > MATERIAL and is thus for use only by the intended recipient. If you > received this in error, please contact the sender and delete the e-mail > and its attachments from all computers. > > - > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
RE: Tomcat Alone or tomcat+IIS/Apache
> From: Peddireddy Srikanth [mailto:[EMAIL PROTECTED] > Subject: Re: Tomcat Alone or tomcat+IIS/Apache > > And they argue that as Tomcat it self runs inside a JVM, which inturn > is a single process all the threads etc wil be simulted ones (and not > the native threads) and hence it will not scale up well under high > loads. More urban myth. As another respondent pointed out, all modern JVMs (i.e., from JRE 1.2 on) use native threads and the underlying OS for thread dispatching. Coupled with thread-local object allocation (available since JRE 1.3), scaling of Tomcat itself is not a problem; it runs happily on our 32-CPU servers as long as the applications themselves have no inherent bottlenecks. You do want to adjust the heap parameters for any serious work, especially on a Windows platform, where the default maximum borders on the miniscule. - Chuck THIS COMMUNICATION MAY CONTAIN CONFIDENTIAL AND/OR OTHERWISE PROPRIETARY MATERIAL and is thus for use only by the intended recipient. If you received this in error, please contact the sender and delete the e-mail and its attachments from all computers. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Tomcat Alone or tomcat+IIS/Apache
We had the same discussion a year ago, as we switched to tomcat 5 and was testing whether we do need apache in front of it. Actually the only advantage for this solution left were apache mods like url-rewriting -> http://mydomain -> http://mydomain/myapp/mypath <- better for some search engines and so on. mod_gzip -> now supported by tomcat directly. But this isn't worth installing an apache. The real problem with serving static content that tomcat has is the thread-model -> one thread per connection is a bit messy if you have http 1.1 (keep-alives) on. But apache has the same issue. So if you have a log of static content to serve I'd strongly suggest you put a squid in front of your tomcats instead of apache (iis is a joke anyway). regards leon On 9/30/05, Mladen Turk <[EMAIL PROTECTED]> wrote: > Peddireddy Srikanth wrote: > > And they argue that as Tomcat it self runs inside a JVM, which inturn > > is a single process all the threads etc wil be simulted ones (and not > > the native threads) and hence it will not scale up well under high > > loads. > > > > Is this argument a valid one or just a misunderstanding?? > > > > The statement for threads is probably valid if you read some > Java 1.1 book. From version 1.2 Java uses platform native threads. > > Regards, > Mladen. > > - > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Tomcat Alone or tomcat+IIS/Apache
Peddireddy Srikanth wrote: And they argue that as Tomcat it self runs inside a JVM, which inturn is a single process all the threads etc wil be simulted ones (and not the native threads) and hence it will not scale up well under high loads. Is this argument a valid one or just a misunderstanding?? The statement for threads is probably valid if you read some Java 1.1 book. From version 1.2 Java uses platform native threads. Regards, Mladen. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Tomcat Alone or tomcat+IIS/Apache
Dear Chuck, your mail is very informative. We r facing a similar issue in our organization :: Tomcat or Apache+tomcat. And the supporters of Apache+tomcat are arguing that as Apache/IIS can make use of native OS (windows inour case) libraries for thread management , memory mangement etc, they will fare well under high loads. And they argue that as Tomcat it self runs inside a JVM, which inturn is a single process all the threads etc wil be simulted ones (and not the native threads) and hence it will not scale up well under high loads. Is this argument a valid one or just a misunderstanding?? thanx for any kind of info in this regard. Regards Srikanth On 9/20/05, Caldarale, Charles R <[EMAIL PROTECTED]> wrote: > > From: Gregg D Bolinger [mailto:[EMAIL PROTECTED] > > Subject: Tomcat Alone or tomcat+IIS/Apache > > > > I know that delivering static content with Apache/IIS is > > preferred. > > Urban myth, based primarily on older Tomcat versions that did not > perform anywhere near as well as the current one. > > > But does that matter if every single request has > > to go to Tomcat because the data is dynamic? > > Think about it: How could adding path length and latency for every > request improve performance? > > > What is the benefit of Tomcat + Apache/IIS on major J2EE apps? > > Job security perhaps? Also increased stress levels, if you enjoy that > sort of thing. Unless there's something specific for httpd or IIS to do > (e.g., poor man's load balancing), simplify your life and leave them > out. > > Check out Peter Lin's performance measurements for just static content, > and you may decide you don't need httpd or IIS for that, either. > http://jakarta.apache.org/tomcat/articles/benchmark_summary.pdf > > - Chuck > > > THIS COMMUNICATION MAY CONTAIN CONFIDENTIAL AND/OR OTHERWISE PROPRIETARY > MATERIAL and is thus for use only by the intended recipient. If you > received this in error, please contact the sender and delete the e-mail > and its attachments from all computers. > > - > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
RE: Tomcat Alone or tomcat+IIS/Apache
> From: Gregg D Bolinger [mailto:[EMAIL PROTECTED] > Subject: Tomcat Alone or tomcat+IIS/Apache > > I know that delivering static content with Apache/IIS is > preferred. Urban myth, based primarily on older Tomcat versions that did not perform anywhere near as well as the current one. > But does that matter if every single request has > to go to Tomcat because the data is dynamic? Think about it: How could adding path length and latency for every request improve performance? > What is the benefit of Tomcat + Apache/IIS on major J2EE apps? Job security perhaps? Also increased stress levels, if you enjoy that sort of thing. Unless there's something specific for httpd or IIS to do (e.g., poor man's load balancing), simplify your life and leave them out. Check out Peter Lin's performance measurements for just static content, and you may decide you don't need httpd or IIS for that, either. http://jakarta.apache.org/tomcat/articles/benchmark_summary.pdf - Chuck THIS COMMUNICATION MAY CONTAIN CONFIDENTIAL AND/OR OTHERWISE PROPRIETARY MATERIAL and is thus for use only by the intended recipient. If you received this in error, please contact the sender and delete the e-mail and its attachments from all computers. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Tomcat Alone or tomcat+IIS/Apache
Well I'm sure you can imagine that if all of your content is dynamic then layering tomcat behind Apache/IIS will only add latency/resources to your requests... nothing significant.. but maybe if your serving up a ton of requests it might be worthwhile to run tomcat standalone. -David Quoting Gregg D Bolinger <[EMAIL PROTECTED]>: I am just curious. I know that delivering static content with Apache/IIS is preferred. But does that matter if every single request has to go to Tomcat because the data is dynamic? Is there some caching that gets involved here? What is the benefit of Tomcat + Apache/IIS on major J2EE apps? Gregg This message was sent using IMP, the Internet Messaging Program. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]