Bugzilla (bugzilla.redhat.com) will be under maintenance for infrastructure upgrades and will not be unavailable on July 31st between 12:30 AM - 05:30 AM UTC. We appreciate your understanding and patience. You can follow status.redhat.com for details.
Bug 1410140 - [RFE] Retry failed "Install dependencies" ansible task when engine/hypervisor deployment fail
Summary: [RFE] Retry failed "Install dependencies" ansible task when engine/hypervisor...
Keywords:
Status: NEW
Alias: None
Product: Red Hat Quickstart Cloud Installer
Classification: Red Hat
Component: Installation - RHEV
Version: 1.1
Hardware: All
OS: All
unspecified
unspecified
Target Milestone: ---
: 1.2
Assignee: John Matthews
QA Contact: Sudhir Mallamprabhakara
Dan Macpherson
URL:
Whiteboard:
: 1405082 (view as bug list)
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2017-01-04 14:57 UTC by Landon LaSmith
Modified: 2019-02-18 22:29 UTC (History)
3 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed:
Target Upstream Version:


Attachments (Terms of Use)

Description Landon LaSmith 2017-01-04 14:57:28 UTC
During deployments of RHV or OSP, the ansible role task for "Install dependencies" will fail due to a single missing package. On most attempts, I have been able to resume the failed tasks (from the bottom up) and complete the deployment.  

Unless there is a way to verify that Satellite has sync'd all of the dependencies for a package prior to kicking off the deployment task, I would recommend attempting the task up to 2 times with some delay between attempts


QCI Media Version: QCI-1.1-RHEL-7-20161215.t.0

Comment 2 John Matthews 2017-01-06 21:15:42 UTC
Landon,

Is this issue an intermittent issue you encounter occasionally with 1.1, assume the problem of installing a package does not happen on every run?

Sounds like the underlying issue is a bug with lazy sync.  

I think your retry suggestion would be appropriate for QCI 1.1 if you are seeing this on a somewhat regular basis, if this is a very rare problem I'd suggest we defer to 1.2.

Thoughts?

Comment 3 John Matthews 2017-01-09 14:34:29 UTC
Spoke with Landon, we decided to align this to 1.2 as it hasn't been showing up often.  Landon hasn't seen it occur since the bug was filed.

If this shows up again please update this BZ with pulp/apache logs from the Satellite.

Comment 4 John Matthews 2017-01-11 16:29:42 UTC
*** Bug 1405082 has been marked as a duplicate of this bug. ***


Note You need to log in before you can comment on or make changes to this bug.