yahoo-eng-team team mailing list archive

[Uggla <2120357@xxxxxxxxxxxxxxxxxx>]

Thank you for submitting this bug report.

Please note that OpenStack Victoria is an End-of-Life (EOL) release and
is no longer actively supported by the OpenStack community. We
prioritize bug fixes and support for currently supported releases, such
as Flamingo and Caracal.

To help us investigate this issue effectively, especially if you
consider upgrading to a supported release, and to ensure your bug report
is complete, please provide the following missing information as per our
bug reporting template:

*   Steps to reproduce: Please provide a chronological list of exact steps, including any openstack client commands, that reliably reproduce the slow API response.
*   Expected result: What is the expected response time for the server group list API for 188 server groups? Or, what behavior should have happened if the issue wasn't present?
*   Environment:
    *   Could you please provide a more exact version of OpenStack Victoria you are running? For example, the output of dpkg -l | grep nova or rpm -qa | grep nova, or git log -1 if from a git source.
    *   Which storage type are you using (e.g., Ceph, LVM)?
    *   Which networking type are you using (e.g., Neutron with OpenVSwitch)?
*   Logs & Configs: Please attach relevant logs, especially nova-api logs from the period when the slow API call was made. sosreport can be very helpful for this.

For your reference, here is the link to the bug reporting template:
https://wiki.openstack.org/wiki/Nova/BugsTeam/BugReportTemplate

For now, we’ll mark this bug as 'Incomplete' due to the use of an
unsupported OpenStack version. If you upgrade to a supported OpenStack
release and the issue persists, please update this bug report with the
requested details and set it back to 'New'.

** 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/2120357

Title:
  Obtain server group list is slow

Status in OpenStack Compute (nova):
  Invalid

Bug description:
  We are using Victoria OpenStack and we found server group list API is
  pretty slow. For a total number of 188 server groups, API returns in
  about 8 seconds.

  After some debugging, it turns out
  https://github.com/openstack/nova/blob/04a65154a96d4d3094895f7b3bdad9f77c7d3e96/nova/api/openstack/compute/server_groups.py#L184
  and
  https://github.com/openstack/nova/blob/04a65154a96d4d3094895f7b3bdad9f77c7d3e96/nova/api/openstack/compute/server_groups.py#L109C23-L109C39
  introduces the high performance penalty. It seems to me that
  server_group_member table is never going to be cleanup event after
  instances got deleted and DB item being archived.

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/2120357/+subscriptions