Out Of Memory for BearerBox
Hi, as i've already seen for people in other posts, i get out of memory for bearerbox. My version is 1.4.0. I've 7 SMSc (emi) connections and 3 modems ( siemens tc35i ). We usually send and receive many thousands of messages per day and, beside a little memory leak, everything goes fine. But at a certain hour of the day, the bearerbox takes all of 2Gb of swap memory in few minutes and the machine freezes. Normally we send two message types , binary or textual, with a maximum length of 160 characters and always with DLR. But at that hour of the day we send a report to few persons. The report length is more than 160 characters and has to be splitted in two or more messages with DLR. I've made some testing and i can reproduce the situation simply sending 20 'long' message in few seconds, here is the message: http://127.0.0.1:13131/cgi-bin/sendsms?username=testpassword=testtext=12345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456780to=%2B393479857204from=4320100validity=1440dlrmask=7dlrurl=http%3A%2F%2Fwww1%2Fprivate%2Fsms.jsp%3Fsmsc-id%3D%25i%26status%3D%25d%26answer%3D%25A%26to%3D%25P%26from%3D%25p%26ts%3D%25t%26idprog%3D1 It has the same syntax of the thousands that we send every day except the length. I think that bearebox has a bug that involves the message split and the delivery report. With older version whe haven't this problem. Mario
Re: Out Of Memory for BearerBox
Did you use checking malloc or configure flag with-defaults=debug Aarno On 12.8.2005, at 10.48, Mario Pandolfi wrote: Hi, as i've already seen for people in other posts, i get out of memory for bearerbox. My version is 1.4.0. I've 7 SMSc (emi) connections and 3 modems ( siemens tc35i ). We usually send and receive many thousands of messages per day and, beside a little memory leak, everything goes fine. But at a certain hour of the day, the bearerbox takes all of 2Gb of swap memory in few minutes and the machine freezes. Normally we send two message types , binary or textual, with a maximum length of 160 characters and always with DLR. But at that hour of the day we send a report to few persons. The report length is more than 160 characters and has to be splitted in two or more messages with DLR. I've made some testing and i can reproduce the situation simply sending 20 'long' message in few seconds, here is the message: http://127.0.0.1:13131/cgi-bin/sendsms? username=testpassword=testtext=1234567890123456789012345678901234567 8901234567890123456789012345678901234567890123456789012345678901234567 8901234567890123456789012345678901234567890123456789012345678901234567 80to=% 2B393479857204from=4320100validity=1440dlrmask=7dlrurl=http%3A% 2F%2Fwww1%2Fprivate%2Fsms.jsp%3Fsmsc-id%3D%25i%26status%3D%25d% 26answer%3D%25A%26to%3D%25P%26from%3D%25p%26ts%3D%25t%26idprog%3D1 It has the same syntax of the thousands that we send every day except the length. I think that bearebox has a bug that involves the message split and the delivery report. With older version whe haven't this problem. Mario
Re: Out Of Memory for BearerBox
I downloaded gateway-1.4.0.tar.gz and compiled it without options. We use SuSE 9.2 pro with kernel 2.6.8-24.17 and have this problem since we installed kannel 1.4 on kernel 2.6. Before this we used kannel 1.2 on 2.4 kernel , we haven't tried 1.4 on 2.4. Ok, Maybe too much update. :-) Mario Aarno Syvänen wrote: Did you use checking malloc or configure flag with-defaults=debug Aarno On 12.8.2005, at 10.48, Mario Pandolfi wrote: Hi, as i've already seen for people in other posts, i get out of memory for bearerbox. My version is 1.4.0. I've 7 SMSc (emi) connections and 3 modems ( siemens tc35i ). We usually send and receive many thousands of messages per day and, beside a little memory leak, everything goes fine. But at a certain hour of the day, the bearerbox takes all of 2Gb of swap memory in few minutes and the machine freezes. Normally we send two message types , binary or textual, with a maximum length of 160 characters and always with DLR. But at that hour of the day we send a report to few persons. The report length is more than 160 characters and has to be splitted in two or more messages with DLR. I've made some testing and i can reproduce the situation simply sending 20 'long' message in few seconds, here is the message: http://127.0.0.1:13131/cgi-bin/sendsms? username=testpassword=testtext=1234567890123456789012345678901234567 8901234567890123456789012345678901234567890123456789012345678901234567 8901234567890123456789012345678901234567890123456789012345678901234567 80to=% 2B393479857204from=4320100validity=1440dlrmask=7dlrurl=http%3A% 2F%2Fwww1%2Fprivate%2Fsms.jsp%3Fsmsc-id%3D%25i%26status%3D%25d% 26answer%3D%25A%26to%3D%25P%26from%3D%25p%26ts%3D%25t%26idprog%3D1 It has the same syntax of the thousands that we send every day except the length. I think that bearebox has a bug that involves the message split and the delivery report. With older version whe haven't this problem. Mario
Re: Out of Memory
Alejandro J. Ramirez wrote: Hello All!! This has happened to me quite a few times now: From message log: Jul 24 04:54:38 movil kernel: Out of Memory: Killed process 6186 (bearerbox). I have the same problem with a RH9.0 - this night at 1AM the gateway stopped and I came to reboot it at 3:00 AM. At 10:00 AM this morning, 7 hours later, same problem - machine locked I'm verifying swaps and I have 2 swap partition mounted for 2Gb J
Re: Out of Memory
Julien Buratto wrote: Alejandro J. Ramirez wrote: Hello All!! This has happened to me quite a few times now: From message log: Jul 24 04:54:38 movil kernel: Out of Memory: Killed process 6186 (bearerbox). I have the same problem with a RH9.0 - this night at 1AM the gateway stopped and I came to reboot it at 3:00 AM. At 10:00 AM this morning, 7 hours later, same problem - machine locked I'm verifying swaps and I have 2 swap partition mounted for 2Gb J I've now addedd 512Mb Ram going to 1Gb to see how it goes like that tonight :) J
Re: Out of Memory
On Jul 26, 2005, at 7:55 PM, Alejandro J. Ramirez wrote: Jul 24 04:54:38 movil kernel: Out of Memory: Killed process 6186 (bearerbox). I would suggest you look at your process list with ps to determine exactly which program it is that gobbles up your memory. With more specifics (like the kind of configuration you are using for that particular program), it might be possible for someone to come up with an answer. Cheers, Pi
Re: Out of Memory
I don't think it's bearerbox's fault. It would involve major memory leaks and I don't think that's the case. It looks like you are not short of memory (360Mb out of 1 Gb is more than enough). Maybe you don't have swap space enabled? Please run top again and look for the a line like this: Swap: 1052248k av, 3512k used, 1048736k free 304424k cached If it's not there or it shows 0k size, it's not activated or at least not working as expected. In that case, when you approach your memory limit, you'll start to experience problems. If swap is ok, then the problem should be elsewere, start with ps auxw | more and try to look who's eating all thos megabytes of RAM. Hope it helps, On 7/26/05, Alejandro J. Ramirez [EMAIL PROTECTED] wrote: Hello All!! This has happened to me quite a few times now: From message log: Jul 24 04:54:38 movil kernel: Out of Memory: Killed process 6186 (bearerbox). After this the whole system hangs, you cannot telnet SSH or even use the console. Only hard reboot works. Top command shows this normally: CPU states: 0.3% user 0.3% system 0.0% nice 0.0% iowait 99.2% idle Mem: 1030892k av, 662820k used, 368072k free, 0k shrd, 77780k buff 357484k actv, 121260k in_d, 8232k in_c I have the feeling that his happens after the bearerbox,sqlbox,smsbox have being running for too long without a reset… What do you guys think? Should I have something in my crontab to reset this services on a daily basis?? Has anyone ran into this kind of problem before?? Thank you!! Alejandro-- Alejandro GuerrieriMagicom http://www.magicom-bcn.net/
RE: Out of Memory
Thanx all. This is mi top output concerning swap space 13:18:28 up 1 day, 21:30, 3 users, load average: 0.01, 0.03, 0.00 60 processes: 58 sleeping, 2 running, 0 zombie, 0 stopped CPU states: 0.0% user 0.8% system 0.0% nice 0.0% iowait 99.2% idle Mem: 1030892k av, 675004k used, 355888k free, 0k shrd, 78452k buff 359980k actv, 129764k in_d, 6740k in_c Swap: 522104k av, 0k used, 522104k free 458496k cached I think its quite alright. As for the memory consuming processes. This are the only ones worth mentioning root 19021 0.1 0.3 327656 3972 ? S 12:45 0:02 /home/alex/gateway-1.4.0/gw/bearerbox -d /home/alex/gateway-1.4 .0/gw/smskannel.conf root 19062 0.0 0.2 39124 2716 ? S 12:45 0:00 /home/alex/gateway-1.4.0/gw/sqlbox -d /home/alex/gateway-1.4.0/ gw/smskannel.conf root 19072 0.0 0.3 80704 3416 ? S 12:45 0:01 /home/alex/gateway-1.4.0/gw/smsbox -d /home/alex/gateway-1.4.0/ gw/smskannel.conf apache 19123 0.1 0.8 18464 8780 ? S 12:57 0:02 /usr/sbin/httpd apache 19131 0.1 0.8 18472 8792 ? S 13:05 0:01 /usr/sbin/httpd mysql 1912 0.6 1.1 15804 11432 ? S Jul24 17:36 /usr/libexec/mysqld --defaults-file=/etc/my.cnf --basedir=/usr --datadir=/var/lib/mysql --user=mysql --pid-file=/var/run/mysqld/mysqld.pid --skip-locking Im running this on a typical Redhat 9.0 installation. Ive installed from sources. Ive had this happen to me prior to the sqlbox installation I upgraded to 1 Gb memory and it was quite for sometime. But now it wants MORE? Do you think it might be wise to maybe install the smsbox on some other server?? Or the the sqlbox maybe? Can this even be done? Ideas welcome, Cheers, Alejandro -Mensaje original- De: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] En nombre de Alejandro Guerrieri Enviado el: Martes, 26 de Julio de 2005 11:31 a.m. Para: users@kannel.org Asunto: Re: Out of Memory I don't think it's bearerbox's fault. It would involve major memory leaks and I don't think that's the case. It looks like you are not short of memory (360Mb out of 1 Gb is more than enough). Maybe you don't have swap space enabled? Please run top again and look for the a line like this: Swap: 1052248k av, 3512k used, 1048736k free 304424k cached If it's not there or it shows 0k size, it's not activated or at least not working as expected. In that case, when you approach your memory limit, you'll start to experience problems. If swap is ok, then the problem should be elsewere, start with ps auxw | more and try to look who's eating all thos megabytes of RAM. Hope it helps, On 7/26/05, Alejandro J. Ramirez [EMAIL PROTECTED] wrote: Hello All!! This has happened to me quite a few times now: From message log: Jul 24 04:54:38 movil kernel: Out of Memory: Killed process 6186 (bearerbox). After this the whole system hangs, you cannot telnet SSH or even use the console. Only hard reboot works. Top command shows this normally: CPU states: 0.3% user 0.3% system 0.0% nice 0.0% iowait 99.2% idle Mem: 1030892k av, 662820k used, 368072k free, 0k shrd, 77780k buff 357484k actv, 121260k in_d, 8232k in_c I have the feeling that his happens after the bearerbox,sqlbox,smsbox have being running for too long without a reset What do you guys think? Should I have something in my crontab to reset this services on a daily basis?? Has anyone ran into this kind of problem before?? Thank you!! Alejandro -- Alejandro Guerrieri Magicom http://www.magicom-bcn.net/
Out of memory!! BEARERBOX
Hello All! My Kannel box is dying on me this is what the Console shows. Ive only got 256 (Im working on it I know I know). Still is there a way to make this process take priority?? (renice it??) May 9 16:39:06 cadatmana05 kernel: Out of Memory: Killed process 2547 (bearerbox). After this the whole system hangs and I cannot even shut it down cleanly Anyone else with this problem?? Alejandro Ramrez
