Hello,
I've been through these dumps but I don't have the experience to learn
anything useful. The only working theory I have is related to the
registrant module. I'll test in the lab with a high number of entries to
see if the problem occurs sooner. In the mean time, is there a more
skilled set of eyes available to look at these memory dumps?
http://storageb.fidelityvoice.net/~jpyle/mem-20130927.log.gz
Thanks,
Jeff
On Fri, Sep 27, 2013 at 10:19 PM, Jeff Pyle jp...@fidelityvoice.com wrote:
Hello,
On Opensips 1.9 from 2013-08-13 I've been running into pkg_memory errors
after a period of uptime. Entires like this appear in the log.
opensips[21368]: ERROR:core:parse_from_header: out of pkg_memory
opensips[21368]: ERROR:uac:restore_uris_reply: failed to find/parse FROM
hdr
opensips[21368]: WARNING:core:fm_malloc: Not enough free memory, will
atempt defragmenation
This time it happened to be something with the From header. Various
options fail; if there's no memory that's understandable.
I'm following the directions from the memory troubleshooting page up to
and including the last line that says if I don't understand the output,
post it here. I don't so it's attached.
Traffic is relatively low with less 10 concurrent calls and less than 1
cps average. Traffic amount doesn't seem to influence the run time before
memory problems.
This instance began to give me trouble a few days after I increased the
number of registrant user entries from 1 to 8. That is the only change I'm
aware of between when it ran for weeks with no problems and days before the
pkg_memory log entries. I have other near-equally configured instances
with only 1 registrant user defined and they continue okay.
Does this log point to anything specific?
Regards,
Jeff
___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel