yahoo-eng-team team mailing list archive

[suibinz <suibin@xxxxxxxx>]

confirmed that this was caused by hosts running short on RAM limit. But
not sure where the limit was set (maybe in Rabbitmq-server?)

** Changed in: nova
       Status: New => Invalid

-- 
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/1255658

Title:
  Error launching an instance: nova scheduler driver: Setting instance
  to ERROR state.

Status in OpenStack Compute (Nova):
  Invalid

Bug description:
  My stack (one all-in-one node + one compute node) had been working fine.  Then it started to have error launching instances with the following observations:
  1) Horizon reports success, but instance status is "Error";
  2) no error/warning from nova-compute.log
  3) nova-scheduler.log has this error: 
  WARNING nova.scheduler.driver [req-ff98dcde-c88b-40e1-85c4-6bb89b627c44 24477163d7ca46a38b9d45360a395d59 8db3509086494bc3a0a5174c26392bb1] [instance: 380fbb79-dbdb-407d-bd89-78afeba8e83d] Setting instance to ERROR state.
  4) "nova-manage service list" shows everything is working properly on both nodes.
  5) there are plenty of diskspace (mounted nfs on /var/lib/nova/instances)
  6) restarted all nova/quantum services but did not help.

  Some suggested to check rabbitmq-plugin for the Disk limit, but there
  was no warning/error on Disk limits.

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/1255658/+subscriptions