Re: EMI: Serious Problem PANIC: Too many concurrent allocations
On Wed, 2002-06-05 at 09:39, Aarno Syvänen wrote: Paul Keogh kirjoittaa tiistaina, 4. kesäkuuta 2002, kello 14:51: Yes, but when this is triggered bb_smscconn_receive () logs the event and returns -1. All the SMSC drivers except HTTP ignore the return code from bb_smscconn_receive (). Therefore, the message is silently dropped from the application and the SMSC point of view. This is IMHO a bad thing and not something you could use in a production environment. I think a better solution would be to; * When possible, map the queue full event to an SMSC protocol error indicating a temporary resource shortage; otherwise fail the message with the most appropriate error code. * Introduce a flow control admin. message to tell the SMS box (and any other clients using the SMS box interface) to stop/start sending messages. The SMS box could in turn signal to the various sendsms applications that a temporary resource shortage event has occurred (HTTP 503 maybe ?) * Use high and low watermark variables instead of maximum-queue-length. This prevents thrashing around the maximum-queue-length value. A sort of SMS hysteresis curve :-). Maximum-queue-length is supposed to prevent crashing caused by too long queues. Congestion control is used to *prevent* long queues. It is, of course, something Kannel can use. Aarno I've been looking at the code and I can't find what I've been looking for :((( On May 3, as I told you, I've sent 200k messages through emi2 (30msg/sec I think). On that day I've tryed my post-xml code. As 100k per post gave me http timeout, I've send 10k at a time, 20 posts. At the end, I've lost 25% of the messages (smsbox logs vs bearerbox logs). Could it be from this code ? I did got store.lock with 30 or 40 MB. As I saw in the code, even if this code is activated, we get a DROPPED in logs, right ? I didn't care much at the time, but there's somewhere a bug looking for us.
RE: EMI: Serious Problem PANIC: Too many concurrent allocations
There may be a possibility on the lack of ram on your machine maybe? On our production machine, there are times during high traffic hours where apache plus kannel was eating up almost everything in terms of ram and cpu. Pushing close to the kind of traffic you are talking about as well. Didn't run into the too many concurrent allocations issue after using a --malloc=native flag since we went live on May 31st. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Bruno David Rodrigues Sent: Wednesday, June 12, 2002 4:23 PM To: Aarno Syv?en Cc: Paul Keogh; [EMAIL PROTECTED] Subject: Re: EMI: Serious Problem PANIC: Too many concurrent allocations On Wed, 2002-06-05 at 09:39, Aarno Syvänen wrote: Paul Keogh kirjoittaa tiistaina, 4. kesäkuuta 2002, kello 14:51: Yes, but when this is triggered bb_smscconn_receive () logs the event and returns -1. All the SMSC drivers except HTTP ignore the return code from bb_smscconn_receive (). Therefore, the message is silently dropped from the application and the SMSC point of view. This is IMHO a bad thing and not something you could use in a production environment. I think a better solution would be to; * When possible, map the queue full event to an SMSC protocol error indicating a temporary resource shortage; otherwise fail the message with the most appropriate error code. * Introduce a flow control admin. message to tell the SMS box (and any other clients using the SMS box interface) to stop/start sending messages. The SMS box could in turn signal to the various sendsms applications that a temporary resource shortage event has occurred (HTTP 503 maybe ?) * Use high and low watermark variables instead of maximum-queue-length. This prevents thrashing around the maximum-queue-length value. A sort of SMS hysteresis curve :-). Maximum-queue-length is supposed to prevent crashing caused by too long queues. Congestion control is used to *prevent* long queues. It is, of course, something Kannel can use. Aarno I've been looking at the code and I can't find what I've been looking for :((( On May 3, as I told you, I've sent 200k messages through emi2 (30msg/sec I think). On that day I've tryed my post-xml code. As 100k per post gave me http timeout, I've send 10k at a time, 20 posts. At the end, I've lost 25% of the messages (smsbox logs vs bearerbox logs). Could it be from this code ? I did got store.lock with 30 or 40 MB. As I saw in the code, even if this code is activated, we get a DROPPED in logs, right ? I didn't care much at the time, but there's somewhere a bug looking for us.
Re: EMI: Serious Problem PANIC: Too many concurrent allocations
Bruno David Rodrigues kirjoittaa keskiviikkona, 12. kesäkuuta 2002, kello 12:22: Yes, it indeed seems that this code can use some rewriting :( Does Paul's suggestion met your big file needs (let us plan better this time). Aarno Paul Keogh kirjoittaa tiistaina, 4. kesäkuuta 2002, kello 14:51: Yes, but when this is triggered bb_smscconn_receive () logs the event and returns -1. All the SMSC drivers except HTTP ignore the return code from bb_smscconn_receive (). Therefore, the message is silently dropped from the application and the SMSC point of view. This is IMHO a bad thing and not something you could use in a production environment. I think a better solution would be to; * When possible, map the queue full event to an SMSC protocol error indicating a temporary resource shortage; otherwise fail the message with the most appropriate error code. * Introduce a flow control admin. message to tell the SMS box (and any other clients using the SMS box interface) to stop/start sending messages. The SMS box could in turn signal to the various sendsms applications that a temporary resource shortage event has occurred (HTTP 503 maybe ?) * Use high and low watermark variables instead of maximum-queue-length. This prevents thrashing around the maximum-queue-length value. A sort of SMS hysteresis curve :-). Maximum-queue-length is supposed to prevent crashing caused by too long queues. Congestion control is used to *prevent* long queues. It is, of course, something Kannel can use. Aarno I've been looking at the code and I can't find what I've been looking for :((( On May 3, as I told you, I've sent 200k messages through emi2 (30msg/sec I think). On that day I've tryed my post-xml code. As 100k per post gave me http timeout, I've send 10k at a time, 20 posts. At the end, I've lost 25% of the messages (smsbox logs vs bearerbox logs). Could it be from this code ? I did got store.lock with 30 or 40 MB. As I saw in the code, even if this code is activated, we get a DROPPED in logs, right ? I didn't care much at the time, but there's somewhere a bug looking for us.
How to change Kannel to show localtime (in detail)
Dear all, Pardon my ignorance (I am kind of new in both Linux and Kannel), but I think GMT datetime issue not only affects the logs but also the timestamp of incoming SMS to the gateway. Can someone guide me on how to change the config to show localtime instead of GMT? I tried the command below but it didn't work for me. My apologies again, I'm sure some of the experts are already groaning to this. :) --Michael Tung Message: 4 Subject: Re: Time.date not right From: Alan McNatty [EMAIL PROTECTED] To: Kannel Dev [EMAIL PROTECTED] Date: 07 Jun 2002 13:03:56 +1200 kannel logs by default in gmt not localtime ... ./config --enable-localtime if you want localtime. On Fri, 2002-06-07 at 06:58, Patrick Mignott wrote: hi all, i see that the time/date in the kannel.log file logging wrong, it's 4 hour faster then the BOIS time why? --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.365 / Virus Database: 202 - Release Date: 5/24/2002
RE: EMI: Serious Problem PANIC: Too many concurrent allocations
On Wed, 2002-06-12 at 10:40, Warren Liu wrote: There may be a possibility on the lack of ram on your machine maybe? On our production machine, there are times during high traffic hours where apache plus kannel was eating up almost everything in terms of ram and cpu. Pushing close to the kind of traffic you are talking about as well. Didn't run into the too many concurrent allocations issue after using a --malloc=native flag since we went live on May 31st. With 1GB ? I don't think so. I recall that smsbox*.log have each and every message but bearerbox_*.log lost almost 25%.
Re: EMI: Serious Problem PANIC: Too many concurrent allocations
Bruno David Rodrigues wrote: I recall that smsbox*.log have each and every message but bearerbox_*.log lost almost 25%. but those messages that have been logged in smsbox.log, but *not* in bearerbox.log have to be queued in the store file, right?! Can you verify this? BTW, I'd like to implement something like a HTTP status page for printing the contents of the store file to an HTTP client. This way we may look into the store file without the need to process it somehow via the command line. Stipe [EMAIL PROTECTED] --- Wapme Systems AG Vogelsanger Weg 80 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are
Re: REF:- SMSC - GSM Modem - Modem Type Auto Detect
Sorry my earlier email was in error and I am looking smsc_at2.c and I do have smsc=at2. I am expecting to see a command like AT+GMM / AT+GMI but cannot see them. Maybe I have an old version - 1.1.6?? Steve - Original Message - From: Andrea Viscovich To: Stephen Rothwell ; [EMAIL PROTECTED] Sent: Wednesday, June 12, 2002 11:29 AM Subject: Re: REF:- SMSC - GSM Modem - Modem Type Auto Detect Look in smsc_at2.c , and use smsc=at2. Andrea - Original Message - From: Stephen Rothwell To: [EMAIL PROTECTED] Sent: Wednesday, June 12, 2002 12:21 PM Subject: REF:- SMSC - GSM Modem - Modem Type Auto Detect Guys, I am looking at the code here in smsc_at.c and cannot seem to see where the manufacturer string and issue are generated ready to be match to the detect-string variables as outlined in the user guide. Is this stuff fully implemented? or am I looking in the wrong place. If it is not done yet I can add the code to do the auto detect but am currently not fully confident of the whole architecture so someone may have to point me in the direction for implementing the hand down of the modem group stuff from the config files. Steve Stephen RothwellTechnical ConsultantConsult HyperionTel: +44 1483 301793Fax: +44 1483 561657www.consult.hyperion.co.uk -where people, networks and money come together...Consult Hyperion-
AT1
On Wed, 2002-06-12 at 11:21, Stephen Rothwell wrote: Guys, I am looking at the code here in smsc_at.c Could we drop AT module on release 1.2 ? Is it there something still lacking in AT2 ? +1 from me. Same question applies to EMI version 1. -- Andreas Fink Fink-Consulting -- Tel: +41-61-6932730 Fax: +41-61-6932729 Mobile: +41-79-2457333 Address: A. Fink, Schwarzwaldallee 16, 4058 Basel, Switzerland E-Mail: [EMAIL PROTECTED] Homepage: http://www.finkconsulting.com -- Something urgent? Try http://www.smsrelay.com/ Nickname afink
Re: AT1
Yep - IMO AT2 is now stable enough (in fact significantly more stable than AT1) to drop AT1. - Original Message - From: Andreas Fink [EMAIL PROTECTED] To: Bruno David Rodrigues [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Wednesday, June 12, 2002 11:50 AM Subject: AT1 On Wed, 2002-06-12 at 11:21, Stephen Rothwell wrote: Guys, I am looking at the code here in smsc_at.c Could we drop AT module on release 1.2 ? Is it there something still lacking in AT2 ? +1 from me. Same question applies to EMI version 1. -- Andreas Fink Fink-Consulting -- Tel: +41-61-6932730 Fax: +41-61-6932729 Mobile: +41-79-2457333 Address: A. Fink, Schwarzwaldallee 16, 4058 Basel, Switzerland E-Mail: [EMAIL PROTECTED] Homepage: http://www.finkconsulting.com -- Something urgent? Try http://www.smsrelay.com/ Nickname afink
Re: AT1
On Wed, 2002-06-12 at 11:50, Andreas Fink wrote: On Wed, 2002-06-12 at 11:21, Stephen Rothwell wrote: Guys, I am looking at the code here in smsc_at.c Could we drop AT module on release 1.2 ? Is it there something still lacking in AT2 ? +1 from me. Same question applies to EMI version 1. But don't emi1 supports X25 and emi2 don't ? (Not that I like x25) -- Andreas Fink Fink-Consulting -- Tel: +41-61-6932730 Fax: +41-61-6932729 Mobile: +41-79-2457333 Address: A. Fink, Schwarzwaldallee 16, 4058 Basel, Switzerland E-Mail: [EMAIL PROTECTED] Homepage: http://www.finkconsulting.com -- Something urgent? Try http://www.smsrelay.com/ Nickname afink
Re: AT1
Yep - IMO AT2 is now stable enough (in fact significantly more stable than AT1) to drop AT1. +1 for dropping smsc_emi.c and smsc_at.c. If we need the code afterwords, it will be in the cvs frozen. Stipe [EMAIL PROTECTED] --- Wapme Systems AG Vogelsanger Weg 80 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are
Re: REF:- SMSC - GSM Modem - Modem Type Auto Detect
On Wed, 2002-06-12 at 11:48, Stephen Rothwell wrote: Sorry my earlier email was in error and I am looking smsc_at2.c and I do have smsc=at2. I am expecting to see a command like AT+GMM / AT+GMI but cannot see them. Maybe I have an old version - 1.1.6?? ATI, line 1876
Dropping EMI1
Same question applies to EMI version 1. But don't emi1 supports X25 and emi2 don't ? (Not that I like x25) there is no native support of X25 anywhere in Kannel so in all cases you have to connect throught tcp/ip to a router which makes X25 out of it. Maybe there is support for serial port connections to a X.28 PAD but I dont think so. -- Andreas Fink Fink-Consulting -- Tel: +41-61-6932730 Fax: +41-61-6932729 Mobile: +41-79-2457333 Address: A. Fink, Schwarzwaldallee 16, 4058 Basel, Switzerland E-Mail: [EMAIL PROTECTED] Homepage: http://www.finkconsulting.com -- Something urgent? Try http://www.smsrelay.com/ Nickname afink
RE: How to change Kannel to show localtime (in detail)
Usually time stamps of SMS messages received by kannel are generated by the SMSC (or sooner) and so Kannel has no influence on them. the SMSC driver should try to convert the time stamp to something meaningful (for example : local timestamps), but it is not always possible, as sometimes the SMSC sends timestamps that are not GMT or the localtime at Kannel location and no indication of what is the correct time zone. -- Oded Arbel m-Wise Mobile Solutions [EMAIL PROTECTED] Mobile: +972-67-340014 Tel: +972-9-9581711 (ext: 116) ::.. The best lessons in life are learned at the worst times. -- Ani DiFranco -Original Message- From: Michael Tung [mailto:[EMAIL PROTECTED]] Sent: Wednesday, June 12, 2002 12:03 PM To: [EMAIL PROTECTED] Subject: How to change Kannel to show localtime (in detail) Dear all, Pardon my ignorance (I am kind of new in both Linux and Kannel), but I think GMT datetime issue not only affects the logs but also the timestamp of incoming SMS to the gateway. Can someone guide me on how to change the config to show localtime instead of GMT? I tried the command below but it didn't work for me. My apologies again, I'm sure some of the experts are already groaning to this. :) --Michael Tung Message: 4 Subject: Re: Time.date not right From: Alan McNatty [EMAIL PROTECTED] To: Kannel Dev [EMAIL PROTECTED] Date: 07 Jun 2002 13:03:56 +1200 kannel logs by default in gmt not localtime ... ./config --enable-localtime if you want localtime. On Fri, 2002-06-07 at 06:58, Patrick Mignott wrote: hi all, i see that the time/date in the kannel.log file logging wrong, it's 4 hour faster then the BOIS time why? --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.365 / Virus Database: 202 - Release Date: 5/24/2002
RE: AT1
+1 to both. -- Oded Arbel m-Wise Mobile Solutions [EMAIL PROTECTED] Mobile: +972-67-340014 Tel: +972-9-9581711 (ext: 116) ::.. Politicians should read science fiction, not westerns and detective stories. -- Arthur C Clarke -Original Message- From: Andreas Fink [mailto:[EMAIL PROTECTED]] Sent: Wednesday, June 12, 2002 12:51 PM To: Bruno David Rodrigues Cc: [EMAIL PROTECTED] Subject: AT1 On Wed, 2002-06-12 at 11:21, Stephen Rothwell wrote: Guys, I am looking at the code here in smsc_at.c Could we drop AT module on release 1.2 ? Is it there something still lacking in AT2 ? +1 from me. Same question applies to EMI version 1. -- Andreas Fink Fink-Consulting -- Tel: +41-61-6932730 Fax: +41-61-6932729 Mobile: +41-79-2457333 Address: A. Fink, Schwarzwaldallee 16, 4058 Basel, Switzerland E-Mail: [EMAIL PROTECTED] Homepage: http://www.finkconsulting.com -- Something urgent? Try http://www.smsrelay.com/ Nickname afink
Message store (Was: EMI: Serious Problem PANIC: Too many concurrent allocations)
Speaking of store files - After really looking into the store logic, it seems my notion of how store worked was all wrong : the elimination of messages that received acks/nacks from storage is done only on store_load, so it's quite possible that a long running Kannel will develop mighty store files which can bring any machine to its proverbial knees. IMO, it's not the right thing to do : message-ack matching should be done on-line, thus preventing issues of too big store files. What do you people think ? -- Oded Arbel m-Wise Mobile Solutions [EMAIL PROTECTED] Mobile: +972-67-340014 Tel: +972-9-9581711 (ext: 116) ::.. Politicians should read science fiction, not westerns and detective stories. -- Arthur C Clarke -Original Message- From: Stipe Tolj [mailto:[EMAIL PROTECTED]] Sent: Wednesday, June 12, 2002 12:36 PM To: Bruno David Rodrigues Cc: [EMAIL PROTECTED]; 'Aarno Syv?en'; 'Paul Keogh'; [EMAIL PROTECTED] Subject: Re: EMI: Serious Problem PANIC: Too many concurrent allocations Bruno David Rodrigues wrote: I recall that smsbox*.log have each and every message but bearerbox_*.log lost almost 25%. but those messages that have been logged in smsbox.log, but *not* in bearerbox.log have to be queued in the store file, right?! Can you verify this? BTW, I'd like to implement something like a HTTP status page for printing the contents of the store file to an HTTP client. This way we may look into the store file without the need to process it somehow via the command line. Stipe [EMAIL PROTECTED] --- Wapme Systems AG Vogelsanger Weg 80 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are
Re: Dropping EMI1
On Wed, 2002-06-12 at 12:19, Andreas Fink wrote: Same question applies to EMI version 1. But don't emi1 supports X25 and emi2 don't ? (Not that I like x25) there is no native support of X25 anywhere in Kannel so in all cases you have to connect throught tcp/ip to a router which makes X25 out of it. Maybe there is support for serial port connections to a X.28 PAD but I dont think so. (FWD to users) Please if someone uses AT or EMI1 module and have some reason to use them instead of AT2 or EMI2, to speak now or shut up forever (How do you say the sentence on marriage ?) :) Paskovsky ? are you still around ? do you use X25 with emi2 ?
Re: EMI: Serious Problem PANIC: Too many concurrent allocations
On Wed, 2002-06-12 at 11:36, Stipe Tolj wrote: Bruno David Rodrigues wrote: I recall that smsbox*.log have each and every message but bearerbox_*.log lost almost 25%. but those messages that have been logged in smsbox.log, but *not* in bearerbox.log have to be queued in the store file, right?! Can you verify this? How can I ? I was injecting 10k on smsbox and they were going out through the other way at 30msg/sec.
Re: EMI: Serious Problem PANIC: Too many concurrent allocations
Bruno David Rodrigues wrote: On Wed, 2002-06-12 at 11:36, Stipe Tolj wrote: Bruno David Rodrigues wrote: I recall that smsbox*.log have each and every message but bearerbox_*.log lost almost 25%. but those messages that have been logged in smsbox.log, but *not* in bearerbox.log have to be queued in the store file, right?! Can you verify this? How can I ? I was injecting 10k on smsbox and they were going out through the other way at 30msg/sec. what I meant was: can you see if the lost messages are still in the kannel.store file? Stipe [EMAIL PROTECTED] --- Wapme Systems AG Vogelsanger Weg 80 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are
