keepalived enters failed state on all RPM family variants

Bug #1549857 reported by Serguei Bezverkhi on 2016-02-25
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Sam Yaple

Bug Description

When deploying with kolla-ansible deploy, the script gets stuck on waiting for VIP address, but the target bare metal server has VIP:port in listen mode. It looks like HAPROXY does not send back arp replies for VIP address.

Steven Dake (sdake) wrote :

centos source and centos binary are missing a key binary introduced in a change to the check script in keepalived.

Changed in kolla:
status: New → Confirmed
importance: Undecided → Critical
milestone: none → mitaka-3
summary: - deployment waiting for VIP address until timed out
+ keepalived enters failed state on all RPM family variants
Sam Yaple (s8m) wrote :

Nothing was introduced. socat has been used since inception. There is no binary missing at all for Centos.

Sam Yaple (s8m) wrote :

This is not a bug, it was old images being used with new playbooks

Changed in kolla:
status: Confirmed → Invalid
Steven Dake (sdake) wrote :

Note the root cause was that this was a multinode install and the pull action was not being run prior to a redeploy. There are docs fixes in the queue to help people avoid this problem in the future.

Changed in kolla:
assignee: nobody → Sam Yaple (s8m)
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers