Bug 1649432

Summary:	cluster getting replicated on subsequent host discoveries on RHV
Product:	Red Hat CloudForms Management Engine	Reporter:	Felix Dewaleyne <fdewaley>
Component:	Providers	Assignee:	Boriso <bodnopoz>
Status:	CLOSED WORKSFORME	QA Contact:	Sudhir Mallamprabhakara <smallamp>
Severity:	low	Docs Contact:	Red Hat CloudForms Documentation <cloudforms-docs>
Priority:	medium
Version:	5.9.4	CC:	dmetzger, fdewaley, istein, jfrey, jhardy, mperina, obarenbo
Target Milestone:	GA
Target Release:	5.9.7
Hardware:	All
OS:	All
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2018-12-11 21:28:05 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	RHEVM	Target Upstream Version:
Embargoed:

Description Felix Dewaleyne 2018-11-13 15:42:42 UTC

Description of problem:
if due to a network issue host discovery gets triggered multiple times, clusters get duplicated in CloudForms

Version-Release number of selected component (if applicable):
5.9.4

How reproducible:
all the time

Steps to Reproduce:
1. set up RHV with several hosts
2.set up cloudforms to the RHV
3.simulate network issues, causing hosts to be discovered again and again

Actual results:
the cluster gets duplicated in cloudforms despite all having the same UUIDs

Expected results:
the cluster isn't added since it is the same.

Additional info:
this is due to a network issue that leads to triggering the discovery of hosts multiple times in RHV.

Comment 3 dmetzger 2018-12-11 21:28:05 UTC

Environmental issue.

Comment 4 Ilanit Stein 2018-12-17 08:53:01 UTC

I failed to reproduce this bug, using the bellow flow.

Versions:
RHV-4.2.8-0.1.el7ev 
CFME-5.9.6.3.20181126231255_be48af6

Flow:
Have a RHV with one host.
Disable fencing on RHV cluster (so that vdsmd service won't be started automatically)
Run on host: 'systemctl stop vdsmd'
On RHV side, the host become 'non responsive'
On CFME side, the host is marked with '?'

Then run on host 'systemctl start vdsmd'
On RHV side, the host became back up.
On CFME, after automatic refresh, the host became up again.


Cluster status on CFME:
The RHV cluster was NOT duplicated.
I checked it under 'Clusters' tab, in provider details page, cluster field,
and in Host details page, cluster field.

I repeated the above flow twice, however did not encounter cluster duplicate.

Maybe this is specific the versions in the reported env, or maybe some additional conditions to have it reproduce.

Comment 5 Martin Perina 2018-12-17 12:30:36 UTC

Based on above changing to CLOSED-WORKSFORME, feel free to reopen if you were able to reproduce with latest CFME/RHV releases and provide reproducing steps and logs