[How-To] Identify a problematic container in case node crashes on starting all containers.

Created:

2016-11-16 13:12:34 UTC

Modified:

2017-04-24 11:43:04 UTC

0

Was this article helpful?


Have more questions?

Submit a request

[How-To] Identify a problematic container in case node crashes on starting all containers.

Applicable to:

  • Web Presence Builder 12.0 for Linux

Symptoms

Node constantly crashes after Virtuozzo service startup, on starting all containers in a bunch.

Cause

The causes for the problem may vary. This article describes an efficient way to identify the faulty container(s) and disable them from starting up.

Resolution

  1. Disable autoboot for all containers. Right after "Virtuozzo Management Service" is started and before containers get the actual time to start, run the commands :

    for /f %i in ('vzlist -Hao veid') do for /f %j in ('vzcfgt get %i autoboot') do echo %i %j >> \\autoboot.txt
    for /f %i in ('vzlist -Hao veid') do vzcfgt set %i autoboot 0
  2. Start containers one by one and find the one that crashes the server :

    for /f %i in ('vzlist -Hao veid') do vzctl start %i
  3. After node reboot, continue starting containers, having isolated the faulty one, detected on the previous step. One of the ways to isolate containers is to rename their private area directory :

    ren D:\\vz\\private\\CTID CTID.broken
  4. Once all bad containers are identified, restore the autoboot settings :

    for /f "tokens=1,2" %i in (\\autoboot.txt) do vzcfgt set %i autoboot %j
Have more questions? Submit a request
Please sign in to leave a comment.