victoria upgrade jobs aren't using content provider passed dlrn hash tag
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
tripleo |
Fix Released
|
Critical
|
Marios Andreou |
Bug Description
At [1] the undercloud-upgrade [2] the minor update and [3] the standalone-upgrade jobs are failing during the upgrade trying to pull containers for image prepare:
* 2020-12-14 22:00:20 | 2020-12-14 22:00:20.851286 | fa163e7f-
The above is from [1]. Using the same logs as an example the hash is passed in the content provider vars with [4]
* registry_
ussuri: 199.204.45.238
victoria: 199.204.45.249
provider_
ussuri: 33e5f8f0b4d9897
victoria: b6ac01208810425
provider_
ussuri: 33e5f8f0b4d9897
victoria: b6ac01208810425
but it is not consumed by the job, looking at the container image prepare params [5]
* tag: 73035ce83cf7ac1
Looks like a nit in [6] - we saw this today because there was a victoria promotion so the value of current-tripleo was updated.
[1] https:/
[2] https:/
[3] https:/
[4] https:/
[5] https:/
[6] https:/
description: | updated |
Changed in tripleo: | |
status: | In Progress → Fix Released |
fix at https:/ /review. opendev. org/c/openstack /tripleo- quickstart/ +/767112 tripleo-quickstart master: Make sure the victoria consumers are using the right dlrn tag