Bug 1392358

Summary:	[RFE] virt-who to have option to disable automated reports when any action is performed on the virt backend
Product:	Red Hat Enterprise Linux 7	Reporter:	Pavel Moravec <pmoravec>
Component:	virt-who	Assignee:	Chris Snyder <csnyder>
Status:	CLOSED WONTFIX	QA Contact:	Eko <hsun>
Severity:	high	Docs Contact:
Priority:	high
Version:	7.3	CC:	aperotti, cdonnell, khowell, ktordeur, yuefliu
Target Milestone:	rc	Keywords:	FutureFeature
Target Release:	---
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:	2017-07-27 14:50:21 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Bug Depends On:
Bug Blocks:	1420851

Description Pavel Moravec 2016-11-07 10:08:55 UTC

Description of problem:
(I follow bugzilla template for this RFE since it is partially a bug as well)

Since some recent change in virt-who (rebase to 0.17?), it does check-in triggered by any action performed on the virt back-end (this can happen with max. frequency of 1 minute). While this change makes actions from back-end propagated to RHSM much faster, it makes VIRTWHO_INTERVAL parameter almost ridiculous on deployments with frequent changes. Such frequent reporting might not be necessary / welcomed in all cases, since it causes redundantly higher load or makes bad user experience:

E.g. in Satellite6, each such report generates a task. Adding or removing a VM every minute or two, Satellite gets an auxiliary&uninteresting (from user point of view) task every minute or two. Therefore, task list is full of such dummy tasks and one has to use filters from beginning to see more important tasks there.

My proposal is to add an option (disabled by default) that prevents triggering of sending the report on every action on the virt back-end.


Version-Release number of selected component (if applicable):
virt-who 0.17-10


How reproducible:
100%


Steps to Reproduce:
1. Configure virt-who with any back-end with any VIRTWHO_INTERVAL
2. Create a VM or delete a VM frequently
3. Monitor frequency of reports sent to RHSM.
4. Have a configuration ability to see jsut reports every VIRTWHO_INTERVAL seconds but not more often.


Actual results:
3. (almost) Every action in 2 trigers a report
4. no such config option


Expected results:
4. there is such config option


Additional info:
See https://bugzilla.redhat.com/show_bug.cgi?id=1369082#c9 .

Comment 1 Eko 2016-11-14 02:51:24 UTC

Hi, 
I think you can try the VIRTWHO_DEBUG option,
if VIRTWHO_DEBUG=1, it will report the hypervisors/guest json info
if VIRTWHO_DEBUG=1, no json info get out

Comment 2 Pavel Moravec 2016-11-14 11:30:09 UTC

There seems to be a user story where the extra virt-who reporting activities reporting to candlepin in Sat6 flooded katello_event_queue in qpidd - katello was not fast enough to consume the messages, causing candlepin performance issues.

Comment 3 Kevin Howell 2017-07-27 14:50:21 UTC

Behavior has changed as of virt-who-0.19-3, virt-who now respects the interval specified regardless of the source of information. There is no option to toggle this behavior. If you feel strongly that there should be, please reopen this RFE.