deployment doesn't fail if ntp servers are unavailable
Bug #1955414 reported by
Alex Schultz
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
tripleo |
Fix Released
|
High
|
Alex Schultz |
Bug Description
Prior to using chrony we would fail if the ntp servers were unavailable.
This was broken by a change from waitsync to makestep https:/
Because NTP is a hard requirement for the deployment to properly work, we need to make sure that we fail if ntp servers are unavailable.
Changed in tripleo: | |
assignee: | nobody → Alex Schultz (alex-schultz) |
status: | New → In Progress |
To post a comment you must log in.
Reviewed: https:/ /review. opendev. org/c/openstack /tripleo- heat-templates/ +/822330 /opendev. org/openstack/ tripleo- heat-templates/ commit/ 8d46c9c381a2487 d759f477948eabe 51d40479b7
Committed: https:/
Submitter: "Zuul (22348)"
Branch: master
commit 8d46c9c381a2487 d759f477948eabe 51d40479b7
Author: Alex Schultz <email address hidden>
Date: Mon Dec 20 07:58:27 2021 -0700
Add chrony waitsync back in
This is a revert of 373a27163e03659 19f707604df5b61 863029d2e4 because
makestep does not fail if the ntp servers are unavailable. This change
adds waitsync back in so it fails if the servers are unavilable but
leaves the makestep in place which should force the sync. If the
servers are unavailable, the waitsync should catch it. If they are
available waitsync should be a relative noop because the makestep would
sync the system. The waitsync has been increased to 5 minutes (from 200
seconds). waitsync will wait 30 times with an interval of 10 seconds
between checks.
Closes-Bug: #1955414 826b2814f72b727 1b43fdca267
Change-Id: I4ac9210f6533e6