yahoo-eng-team team mailing list archive

["Markus Zoeller \(markus_z\)" <mzoeller@xxxxxxxxxxxxxxxxxx>]

This is an automated cleanup. This bug report has been closed because it
is older than 18 months and there is no open code change to fix this.
After this time it is unlikely that the circumstances which lead to
the observed issue can be reproduced.

If you can reproduce the bug, please:
* reopen the bug report (set to status "New")
* AND add the detailed steps to reproduce the issue (if applicable)
* AND leave a comment "CONFIRMED FOR: <RELEASE_NAME>"
  Only still supported release names are valid (LIBERTY, MITAKA, OCATA, NEWTON).
  Valid example: CONFIRMED FOR: LIBERTY


** Changed in: nova
   Importance: High => Undecided

** Changed in: nova
       Status: Confirmed => Expired

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to OpenStack Compute (nova).
https://bugs.launchpad.net/bugs/1388213

Title:
  libvirt: Possible to crash nova compute node via deletion of a
  resizing / rebooting instance (timing bug)

Status in OpenStack Compute (nova):
  Expired

Bug description:
  NOTE: tests run against devstack install on Ubuntu 14.04
  tests conducted via randomized testing, no definitive test case produced yet, but have repeated this several times.

  Using libvirtd (libvirt) 1.2.2

  It appears that deleting a resizing instance can cause the nova compute node to crash with the following error / traceback:
  (more detailed output below)
  screen-n-cpu.2014-10-31-081111.log:354289:2014-10-31 11:01:58.093 TRACE oslo.messaging.rpc.dispatcher   File "/usr/lib/python2.7/dist-packages/libvirt.py", line 896, in createWithFlags
  screen-n-cpu.2014-10-31-081111.log:354290:2014-10-31 11:01:58.093 TRACE oslo.messaging.rpc.dispatcher     if ret == -1: raise libvirtError ('virDomainCreateWithFlags() failed', dom=self)
  screen-n-cpu.2014-10-31-081111.log:354291:2014-10-31 11:01:58.093 TRACE oslo.messaging.rpc.dispatcher libvirtError: Domain not found: no domain with matching uuid '2aadc976-951e-47d6-bb20-9e071a6a89a9' (instance-00000360)

  When this is triggered, the compute node will crash and all instances will end up in ERROR state.
  Have not perfected the timing parameters, but will provide output from two runs below.

  Also noting that this ends up with <instance_id>_resize and
  <instance_id>_del directories in /opt/stack/data/nova/instances on
  these failures.

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/1388213/+subscriptions