Bug 1463808

Summary:	Stop calling back to candlepin for qpid messages
Product:	Red Hat Satellite	Reporter:	Satellite Program <pm-sat>
Component:	Subscription Management	Assignee:	Justin Sherrill <jsherril>
Status:	CLOSED ERRATA	QA Contact:	Radovan Drazny <rdrazny>
Severity:	high	Docs Contact:
Priority:	high
Version:	6.2.9	CC:	bbuckingham, bkearney, cduryee, jcallaha, jsherril, mmccune, pmoravec, rdrazny
Target Milestone:	Unspecified	Keywords:	Performance, PrioBumpField, Triaged
Target Release:	Unused
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	tfm-rubygem-katello-3.4.0	Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:	1431783	Environment:
Last Closed:	2018-02-21 17:30:30 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Comment 2 Radovan Drazny 2017-10-16 08:09:30 UTC

Tested on Satellite 6.3 Snap 18. Ran subscription of 50 cloned virtual machines in sequence. 

# Before the registrations started:
$ grep "Request" /var/log/candlepin/candlepin.log | wc -l
 1

# After the registration of 50 machines finished:
$ grep "Request" /var/log/candlepin/candlepin.log | wc -l
 867

According to the initial report, more than 17 request per machine/subscription seems to be pretty excessive.

Comment 3 Satellite Program 2017-10-16 10:08:37 UTC

Upstream bug assigned to jsherril

Comment 4 Justin Sherrill 2017-10-16 13:39:05 UTC

Radovan,   can you provide your candlepin.log so we can see exactly what was going on?  This is strange as this was verified for 6.2.z and i don't see how it would have broken in 6.3.

Comment 5 Radovan Drazny 2017-10-16 14:16:31 UTC

(In reply to Justin Sherrill from comment #4)
> Radovan,   can you provide your candlepin.log so we can see exactly what was
> going on?  This is strange as this was verified for 6.2.z and i don't see
> how it would have broken in 6.3.

That Sat 6.3 instance is gone, unfortunately. I can try to reproduce the behaviour again, I still have the base client machine used as the clone-source  and registration scripts. I will run the process overnight, and will let you know the result tomorrow.

Comment 10 Radovan Drazny 2017-10-24 13:08:19 UTC

Ok, I did tests you recommended in comment #8 and in BZ1431783 comment #11. Comparing number of Request calls for 50 hosts between 6.2 and 6.3, there was massive decrease for the "Subs attached => Subs removed" path (2020 less calls on 6.3), and less substantial, but still noticeable decrease for the "No subs attached => Subs attached" (465 less calls for 6.3). 
Scenarios A and B from BZ1431783 comment #11 do pass as well.

Comment 11 Bryan Kearney 2018-02-21 17:30:30 UTC

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/RHSA-2018:0336