2 Aug
2019
2 Aug
'19
9:30 a.m.
Hello Andrew,
I compared the 0 calls and 300 calls dump, indeed there is not that much of a difference.
688 bytes consumed more, they were allocated in pv_cache_add. Can you try another 300
calls and see if there is the same increase?
Cheers,
Henning
-root 9302 0.0 5.4 329264 26848 ? Sl 21:18 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
+root 9302 0.1 6.9 330460 33904 ? Sl 21:18 0:01 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
INFO: <core> [core/cfg/cfg_ctx.c:595]: cfg_set_now(): core.mem_dump_pkg has been
changed to 9302
ALERT: <core> [core/pt.c:548]: mem_dump_pkg_cb(): Memory status (pkg) of process
9302:
ALERT: qm_status: (0x7ff63029f010):
ALERT: qm_status: heap size= 33554432
-ALERT: qm_status: used= 485216, used+overhead=815112, free=32739320
-ALERT: qm_status: max used (+overhead)= 821904
+ALERT: qm_status: used= 485904, used+overhead=826096, free=32728336
+ALERT: qm_status: max used (+overhead)= 838568
ALERT: qm_status: dumping all alloc'ed. fragments:
ALERT: qm_status: 0. N address=0x7ff6302d86d8 frag=0x7ff6302d86a0 size=1024
used=1
ALERT: qm_status: alloc'd from core: core/str_hash.h: str_hash_alloc(59)
@@ -2599,23 +2599,49 @@
ALERT: qm_status: 872. N address=0x7ff630365280 frag=0x7ff630365248 size=32 used=1
ALERT: qm_status: alloc'd from core: core/pvapi.c: pv_parse_format(1152)
ALERT: qm_status: start check=f0f0f0f0, end check= c0c0c0c0, abcdefed
-ALERT: qm_status: 886. N address=0x7ff6303662c8 frag=0x7ff630366290 size=128 used=1
+ALERT: qm_status: 890. N address=0x7ff630365ff8 frag=0x7ff630365fc0 size=144 used=1
+ALERT: qm_status: alloc'd from core: core/pvapi.c: pv_cache_add(347)
+ALERT: qm_status: start check=f0f0f0f0, end check= c0c0c0c0, abcdefed
+ALERT: qm_status: 894. N address=0x7ff6303662c8 frag=0x7ff630366290 size=128 used=1
+ALERT: qm_status: alloc'd from core: core/pvapi.c: pv_cache_add(347)
+ALERT: qm_status: start check=f0f0f0f0, end check= c0c0c0c0, abcdefed
+ALERT: qm_status: 919. N address=0x7ff630367550 frag=0x7ff630367518 size=136 used=1
+ALERT: qm_status: alloc'd from core: core/pvapi.c: pv_cache_add(347)
+ALERT: qm_status: start check=f0f0f0f0, end check= c0c0c0c0, abcdefed
+ALERT: qm_status: 922. N address=0x7ff630367e40 frag=0x7ff630367e08 size=136 used=1
+ALERT: qm_status: alloc'd from core: core/pvapi.c: pv_cache_add(347)
+ALERT: qm_status: start check=f0f0f0f0, end check= c0c0c0c0, abcdefed
+ALERT: qm_status: 932. N address=0x7ff6303684c8 frag=0x7ff630368490 size=128 used=1
+ALERT: qm_status: alloc'd from core: core/pvapi.c: pv_cache_add(347)
+ALERT: qm_status: start check=f0f0f0f0, end check= c0c0c0c0, abcdefed
+ALERT: qm_status: 975. N address=0x7ff63036a0a0 frag=0x7ff63036a068 size=144 used=1
ALERT: qm_status: alloc'd from core: core/pvapi.c: pv_cache_add(347)
ALERT: qm_status: start check=f0f0f0f0, end check= c0c0c0c0, abcdefed
Am 02.08.19 um 14:16 schrieb Andrew White:
Hi Henning!
Thanks so much!
I spun up a dev machine with the same specs, Kamailio version and config. On each one,
I’ve done a ps of the single receiver/worker process, as well as a full dump.
0 calls:
https://paste.ubuntu.com/p/Vdrqkfy9Cm/
100 calls:
https://paste.ubuntu.com/p/d765GRqKmx/
200 calls:
https://paste.ubuntu.com/p/rB2rzgZN7S/
300 calls:
https://paste.ubuntu.com/p/mK4gjJ4tcP/
Interestingly there doesn’t appear to be a major difference from 100-300, which seems odd
to me. I feel my dev server isn’t giving me results mirrored with production maybe.
Please let me know what you think if you can!
Thanks,
Andrew
On 2 Aug 2019, at 6:19 am, Henning Westerholt
<hw@skalatan.de<mailto:hw@skalatan.de>> wrote:
Hi Andrew,
answer inline below
Am 01.08.19 um 15:34 schrieb Andrew White:
Thanks Daniel, you’re fantastic!
I have 4 children/workers configured with -m 128 -M 32. The machine in question has 512MB
of memory, 1 core and 1GB swap on an SSD.
I restarted Kamailio with memlog=1 and I’ve been sending batches of 30 calls in. I’ve
noticed 4 of the 13 Kamailio processes going up in memory after each batch, which I
suspect to be the primary children/workers. Immediately post restart:
root 28531 0.7 5.5 329368 27196 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
root 28532 0.6 4.9 329368 24528 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
root 28533 0.6 5.5 329368 27244 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
root 28534 0.7 5.4 329368 26788 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
After about 90 calls:
root 28531 0.0 6.7 330688 32948 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
root 28532 0.0 6.5 330560 32264 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
root 28533 0.0 6.5 330556 32272 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
root 28534 0.0 6.6 330564 32592 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
None of the other 9 Kamailio processes are increasing at all.
I ran corex.pkg_summary against one of them and got the following dump:
https://paste.ubuntu.com/p/SqTF3K5knK/
I can see a lot of allocation to pvapi.c, does this indicate maybe I’m setting PVs that
need to be unset?
Here’s another after another 60 calls:
https://paste.ubuntu.com/p/9WQXqZtfT2/
There is not that much difference between the two runs (diff):
-journal: Suppressed 19796 messages from /system.slice/kamailio.service
+journal: Suppressed 1772 messages from /system.slice/kamailio.service
ALERT: <core> [core/pt.c:548]: mem_dump_pkg_cb(): Memory status (pkg) of process
28531:
ALERT: qm_status: (0x7f0fc3a2a010):
ALERT: qm_status: heap size= 33554432
-ALERT: qm_status: used= 538712, used+overhead=883168, free=32671264
-ALERT: qm_status: max used (+overhead)= 893040
+ALERT: qm_status: used= 590552, used+overhead=939064, free=32615368
+ALERT: qm_status: max used (+overhead)= 948112
ALERT: qm_status: dumping all alloc'ed. fragments:
ALERT: qm_status: 0. N address=0x7f0fc3a636d8 frag=0x7f0fc3a636a0 size=1024
used=1
ALERT: qm_status: alloc'd from core: core/str_hash.h: str_hash_alloc(59)
The second run shows about 50k more memory usage, but from the dumped allocations they are
identical.
Two suggestions:
- your logging system suppresses actually a lot of debugging output from Kamailio (have a
look to the first lines each). Check if its necessary to increase the respective
configuration variable in your logging conf to get all log messages.
- If you are able to execute this test on a test plattform, configure Kamailio to just use
one worker child each, to spot an error easier. Then execute 100 calls, and count the
memory increase, another 100 calls and count. Then you can estimate how much bytes you
actually leak per call. Try another memory dump and see if you spot something. Share your
results again on the list.
root 28531 0.0 6.9 330820 33928 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
root 28532 0.0 6.7 330692 33352 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
root 28533 0.0 6.7 330688 33280 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
root 28534 0.0 6.7 330696 33192 ? Sl 22:48 0:00 /usr/local/sbin/kamailio
-DD -P /var/run/kamailio/kamailio.pid -f /etc/kamailio/kamailio.cfg -m 128 -M 32
The only changes I’ve made on this config over the last couple of weeks (since I saw this
issue) is removing the dispatcher module and adding in a small function in app_ruby (which
I already use) to query redis (which I also already use from app_ruby and make a heap of
queries per call) for some values and write $du manually. I also added in the topoh
module.
It also makes a lot of sense to me to monitor the individual processes rather than the
aggregate. Is there a way to identify simply from bash what processes are workers
programmatically? I’d like to monitor just those individually in my monitoring.
Have a look to the output of corex.ps or core.psx - the worker childs are named
"receiver".
Cheers,
Henning
Thanks!
Andrew
On 1 Aug 2019, at 8:24 pm, Daniel-Constantin Mierla
<miconda@gmail.com<mailto:miconda@gmail.com>> wrote:
Hello,
if it is pkg, then you have to see which process is increasing the use of memory, because
it is private memory, specific for each process. The sum is an indicator, but the
debugging has to be done for a specific process/pid.
Once you indentify a process that is leaking pkg, execute the rpc command:
- https://www.kamailio.org/docs/modules/devel/modules/corex.html#corex.rpc.pk…
When that process is doing some runtime work (e.g., handling of a sip message), the syslog
will get a summary with used pkg chunks. Send those log messages here for analysis. You
have to set memlog core parameter to a value smaller than debug.
Cheers,
Daniel
On 01.08.19 03:43, Andrew White wrote:
Hi all,
I had a Kamailio crash the other day, and some debugging showed I ran out of PKG memory.
Since then I’ve run a simple bash script to compile the amount of memory used by all child
processes, effective /usr/local/sbin/kamcmd pkg.stats | grep real_used summed together.
I’ve graphed out the data, and there’s a clear growth of PKG memory going on, mostly
increasing during our busier daytime hours.
https://i.imgur.com/UTzx2k1.png
Based on this, I suspect either a module loaded or something within my app_ruby conf is
leaking memory.
I’ve been reading through https://www.kamailio.org/wiki/tutorials/troubleshooting/memory,
but I’m a bit nervous, as I’m not really a C/deep memory type of guy. I can see a GDB
script I can attach to Kamailio, but is that going to use significant resources to run or
impact the running process? Is there a newer/better/alternative way to do this, and to
help me break this down?
Thanks!
Andrew
_______________________________________________
Kamailio (SER) - Users Mailing List
sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>
https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users
--
Daniel-Constantin Mierla -- www.asipto.com<http://www.asipto.com/>
www.twitter.com/miconda<http://www.twitter.com/miconda> --
www.linkedin.com/in/miconda<http://www.linkedin.com/in/miconda>
_______________________________________________
Kamailio (SER) - Users Mailing List
sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>
https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users
--
Henning Westerholt - https://skalatan.de/blog/
Kamailio services - https://skalatan.de/services
--
Henning Westerholt - https://skalatan.de/blog/
Kamailio services - https://skalatan.de/services
