DescriptionFabrizio Soppelsa
2020-03-30 07:34:58 UTC
Description of problem:
When a RHOSP overcloud deploy or update runs, in presence of ceph nodes, the ceph-ansible part seems to take a very long time while apparently is in pause.
Business impact:
overcloud deploy increased time to completion (hours slower).
Version-Release number of selected component (if applicable):
3.3
How reproducible:
overcloud deploy
Actual results:
Sample run:
--------------------------------------------
2020-03-27 01:19:21,595 p=31193 u=mistral | INSTALLER STATUS ***************************************************************
2020-03-27 01:19:21,598 p=31193 u=mistral | Install Ceph Monitor : Complete (0:03:25)
2020-03-27 01:19:21,599 p=31193 u=mistral | Install Ceph Manager : Complete (0:02:03)
2020-03-27 01:19:21,599 p=31193 u=mistral | Install Ceph OSD : Complete (0:02:58)
2020-03-27 01:19:21,599 p=31193 u=mistral | Install Ceph RGW : Complete (0:01:52)
2020-03-27 01:19:21,599 p=31193 u=mistral | Install Ceph Client : Complete (1:14:17)
2020-03-27 01:19:21,600 p=31193 u=mistral | Friday 27 March 2020 01:19:21 +0100 (0:00:00.176) 2:29:00.288 **********
2020-03-27 01:19:21,600 p=31193 u=mistral | ===============================================================================
--------------------------------------------
Expected results:
Understand if this depends on other components. Please analyze the provided data and suggest linked upstream/github bugs, if it's ceph-ansible, if this is fixed already, if we can ask for a hotfix.
Some tunings applied (i.e. increased FORK to 100), but not helping definitely.
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.
For information on the advisory, and where to find the updated
files, follow the link below.
If the solution does not work for you, open a new bug report.
https://access.redhat.com/errata/RHBA-2020:2488