[SR-Users] memory problem

[beer Ll]

HI Phil

this is a part of output of

kamcmd pkg.stats


~# kamcmd pkg.stats
{
    entry: 0
    pid: 3515
    rank: 0
    used: 312260
    free: 33167684
    real_used: 386748
    total_size: 33554432
    total_frags: 16
}
{
    entry: 1
    pid: 3516
    rank: 1
    used: 320776
    free: 33158104
    real_used: 396328
    total_size: 33554432
    total_frags: 40
}
{
    entry: 2
    pid: 3517
    rank: 2
    used: 320692
    free: 33158776
    real_used: 395656
    total_size: 33554432
    total_frags: 20
}
.....



when I run for example


kamcmd cfg.set_now_int core mem_dump_pkg 3516


the output is only this:


/usr/sbin/kamailio[3531]: INFO: <core> [cfg/cfg_ctx.c:613]: cfg_set_now():
INFO: cfg_set_now(): core.mem_dump_pkg has been changed to 3516





The db postgres module errors like this


ERROR: db_postgres [km_dbase.c:537]: db_postgres_store_result(): invalid
query, execution aborted

ERROR: db_postgres [km_dbase.c:538]: db_postgres_store_result(): driver
error: PGRES_FATAL_ERROR, ERROR:


Can it causes memory problem ?













On Mon, Sep 26, 2016 at 1:07 PM, Phil Lavin <phil.lavin at cloudcall.com>
wrote:

> This looks like package memory. Can you dump the fragments and send the
> output?
>
>
>
> kamcmd pkg.stats
>
>
>
> This will show you memory usage, for each Kamailio process.
>
>
>
> kamcmd cfg.set_now_int core mem_dump_pkg PID
>
> kamcmd cfg.set_now_int core mem_dump_pkg 1
>
>
>
> This will dump the memory fragments to syslog. You’ll need to replace PID
> with the process ID of one of your affected Kamailio processes, as taken
> from the pkg.stats output.
>
>
>
>
>
> *From:* sr-users [mailto:sr-users-bounces at lists.sip-router.org] *On
> Behalf Of *beer Ll
> *Sent:* 26 September 2016 11:56
> *To:* sr-users at lists.sip-router.org
> *Subject:* [SR-Users] memory problem
>
>
>
> hi all
>
> I have kamailio 4.2.0 on Debian Jessie
>
> the parameters on startup of kamailio are :
>
> SHM_MEMORY=512
> PKG_MEMORY=32
>
> the server is dedicated to the kamailio SIP service with authentication
> and accounting  on a separated db server
>
> the server process about 30/50  call setups per second
>
>
> everythings works good but after 5/6 days the processes lock and I
> received a lot of error log like these:
>
> <core> [rvalue.c:1710]: rval_str_add2(): rv eval out of memory
> <core> [msg_translator.c:362]: received_builder(): out of memory
> <core> [msg_translator.c:2717]: via_builder(): out of memory
> ...
> <core> [msg_translator.c:2423]: build_res_buf_from_sip_req(): out of
> memory; needs 514
> ...
>  <core> [data_lump.c:172]: insert_new_lump_before(): out of memory
> acc [acc_extra.c:267]: extra2strar(): extra2strar: out of memory.
>
>
> after a kamailio reboot evrethings works again for 5/6 days.
>
>
> what can I do to resolve?
>
> _______________________________________________
> SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list
> sr-users at lists.sip-router.org
> http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sip-router.org/pipermail/sr-users/attachments/20160928/439624d9/attachment.html>