why g-wan loops on loading handler scripts and csp script each day at midnight?

Question

I have a strange behavior on my g-wan server: each day at midnight g-wan loops on loading scripts. This is what I see in gwan.log:

[Tue Apr 09 00:00:00 2013 GMT] memory footprint: 1.47 MiB.
[Tue Apr 09 00:00:00 2013 GMT] Host /var/www/gwan/0.0.0.0_8082/#0.0.0.0
[Tue Apr 09 00:00:00 2013 GMT] log files enabled
[Tue Apr 09 00:00:00 2013 GMT] loaded main.c                         39.13 KiB   MD5:15795d7c-42184ef2-c8075784-a3aa84aa
[Tue Apr 09 00:00:00 2013 GMT] loaded process_kv.c                   44.44 KiB   MD5:349b8978-bbebb4eb-120c6f1a-7d06f98e
[Tue Apr 09 00:00:00 2013 GMT] loaded Connection Handler main.c      18.71 KiB   MD5:f624bc05-f51507c3-61b20c9c-ecfe9e19
[Tue Apr 09 00:00:00 2013 GMT] Host /var/www/gwan/0.0.0.0_8083/#0.0.0.0
[Tue Apr 09 00:00:00 2013 GMT] log files enabled
[Tue Apr 09 00:00:00 2013 GMT] loaded main.c                         39.13 KiB   MD5:15795d7c-42184ef2-c8075784-a3aa84aa
[Tue Apr 09 00:00:00 2013 GMT] loaded process_kv.c                   44.44 KiB   MD5:349b8978-bbebb4eb-120c6f1a-7d06f98e
[Tue Apr 09 00:00:00 2013 GMT] loaded Connection Handler main.c      18.71 KiB   MD5:525aa623-2728dd50-0e67a6ad-6763a30b
[Tue Apr 09 00:00:00 2013 GMT] memory footprint: 5.45 MiB.
[Tue Apr 09 00:00:01 2013 GMT] memory footprint: 1.47 MiB.
[Tue Apr 09 00:00:01 2013 GMT] Host /var/www/gwan/0.0.0.0_8082/#0.0.0.0
[Tue Apr 09 00:00:01 2013 GMT] log files enabled
[Tue Apr 09 00:00:01 2013 GMT] loaded main.c                         39.13 KiB   MD5:15795d7c-42184ef2-c8075784-a3aa84aa
[Tue Apr 09 00:00:01 2013 GMT] loaded process_kv.c                   44.44 KiB   MD5:349b8978-bbebb4eb-120c6f1a-7d06f98e
[Tue Apr 09 00:00:01 2013 GMT] loaded Connection Handler main.c      18.71 KiB   MD5:f624bc05-f51507c3-61b20c9c-ecfe9e19
[Tue Apr 09 00:00:01 2013 GMT] Host /var/www/gwan/0.0.0.0_8083/#0.0.0.0
[Tue Apr 09 00:00:01 2013 GMT] log files enabled
[Tue Apr 09 00:00:01 2013 GMT] loaded main.c                         39.13 KiB   MD5:15795d7c-42184ef2-c8075784-a3aa84aa
[Tue Apr 09 00:00:01 2013 GMT] loaded process_kv.c                   44.44 KiB   MD5:349b8978-bbebb4eb-120c6f1a-7d06f98e
[Tue Apr 09 00:00:01 2013 GMT] loaded Connection Handler main.c      18.71 KiB   MD5:525aa623-2728dd50-0e67a6ad-6763a30b
....
....
[Tue Apr 09 00:02:01 2013 GMT] memory footprint: 3.22 MiB.
[Tue Apr 09 00:02:01 2013 GMT] Host /var/www/gwan/0.0.0.0_8082/#0.0.0.0
[Tue Apr 09 00:02:01 2013 GMT] log files enabled
[Tue Apr 09 00:02:01 2013 GMT] loaded main.c                         39.13 KiB   MD5:15795d7c-42184ef2-c8075784-a3aa84aa
[Tue Apr 09 00:02:01 2013 GMT] loaded process_kv.c                   44.44 KiB   MD5:349b8978-bbebb4eb-120c6f1a-7d06f98e
[Tue Apr 09 00:02:01 2013 GMT] loaded Connection Handler main.c      18.71 KiB   MD5:f624bc05-f51507c3-61b20c9c-ecfe9e19
[Tue Apr 09 00:02:01 2013 GMT] Host /var/www/gwan/0.0.0.0_8083/#0.0.0.0
[Tue Apr 09 00:02:01 2013 GMT] log files enabled
[Tue Apr 09 00:02:01 2013 GMT] loaded main.c                         39.13 KiB   MD5:15795d7c-42184ef2-c8075784-a3aa84aa
[Tue Apr 09 00:02:01 2013 GMT] loaded process_kv.c                   44.44 KiB   MD5:349b8978-bbebb4eb-120c6f1a-7d06f98e
[Tue Apr 09 00:02:01 2013 GMT] loaded Connection Handler main.c      18.71 KiB   MD5:525aa623-2728dd50-0e67a6ad-6763a30b
[Tue Apr 09 00:02:01 2013 GMT] memory footprint: 7.13 MiB.
[Tue Apr 09 00:02:11 2013 GMT] ------------------------------------------------
[Tue Apr 09 00:02:11 2013 GMT] G-WAN 4.3.14 64-bit (Mar 14 2013 07:33:12)
[Tue Apr 09 00:02:11 2013 GMT] ------------------------------------------------
[Tue Apr 09 00:02:11 2013 GMT] Local Time: Tue, 09 Apr 2013 02:02:11 GMT+2
....

So during 2 minutes g-wan loops and after g-wan is started normally by monit I believe. Do you have an explanation and a workaround?

I recently change my web hosting provider and from now I use a VM under Microsoft Hyper-V. The problem remains: G-Wan is restarted by Monit which logged the following events:

Wed, 16 Oct 2013 02:02:12: process PID changed from 2171 to 20631
Wed, 16 Oct 2013 02:02:13: process PPID changed from 1 to 2171
Wed, 16 Oct 2013 02:02:43: process PID changed from 20631 to 20975
Wed, 16 Oct 2013 02:03:13: process PID changed from 20975 to 21318
Wed, 16 Oct 2013 02:03:44: process PID changed from 21318 to 21573
Wed, 16 Oct 2013 02:04:14: process PID changed from 21573 to 2171
Wed, 16 Oct 2013 02:04:14: process PPID changed from 2171 to 1
Wed, 16 Oct 2013 02:33:06: process is not running
Wed, 16 Oct 2013 02:33:37: process is running with pid 776

It happens every day even on my new VM with another hypervisor.

I am not the only one with this behavior. Do you have an explanation and a workaround?

Thank you Jerome

same thing here ... and without monit ... GWAN just stop (or crash) — solisoft, Apr 09 '13 at 07:31
Even when I deactivate all the logs G-wan restart at midnight and loops. Any idea? — user1606908, Apr 10 '13 at 07:03
I tried on : archlinux, ubuntu & opensuse, all 64bits version ... all crash at 00h00 — solisoft, Apr 12 '13 at 05:50

Gil · Answer 1 · 2013-08-25T06:11:23.737

0

The loop means that the child process is not responding, hence the parent killing the old child and restarting a new child.

Could you email us the whole log files (trace and gwan.log)? Your partial post is missing the the informative parts.

UPDATE

That new problem for 4-years old code that worked fine so far is a platform issue, for which we have found a workaround, to be published with the next release in a few weeks.

edited Aug 25 '13 at 06:11

answered Apr 10 '13 at 07:21

Gil

3,279
1
15
25

can you tell me on which email I can send the log files? – user1606908 Apr 10 '13 at 11:48
Some platforms (notably hypervisors) seem to be the cause of that issue. Other users are unaffected. – Gil Apr 24 '13 at 06:09

why g-wan loops on loading handler scripts and csp script each day at midnight?

1 Answers1