Bug 1565446

Summary:	Lack of builder resources on amd64 and i386
Product:	[Community] Copr	Reporter:	Marc Dequènes (Duck) <duck>
Component:	backend	Assignee:	clime
Status:	CLOSED CURRENTRELEASE	QA Contact:
Severity:	medium	Docs Contact:
Priority:	unspecified
Version:	unspecified	CC:	clime, praiskup
Target Milestone:	---
Target Release:	---
Hardware:	x86_64
OS:	All
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2018-04-14 08:40:26 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:
Embargoed:

Description Marc Dequènes (Duck) 2018-04-10 03:25:49 UTC

Description of problem:
Building `needrestart` usually took 2-8 minutes overall on various distro/arch combination. Yesterday build #739049 took one hour, and today build #739155 after three hours is still blocked waiting for builder's availability.

Maybe it's a special time/event and this is going to be transient. I see long running builds in the status page so that could explain. Nevertheless it seems the number of builders for these very popular architectures is scarce.

Please note I'm not requesting special treatment. If this is going to be very temporary, then feel free to close this BR. If there's a real problem and more resources are needed, then we (OSAS) could vouch for this service's usefulness. I wonder if we might also be able to help more directly.

Comment 1 clime 2018-04-10 07:47:42 UTC

Thank you. The problem is that COPR currently operates on an old OpenStack cloud infrastructure where the Cloud controller repeteadly suffers from memory leaks, which causes lack of available builders. The plan is to move it to a new platform in the upcoming weeks. I hope we can do it fast. Meanwhile, we somehow need to maintain and live with the current platform.

Comment 2 Marc Dequènes (Duck) 2018-04-10 08:53:00 UTC

Thanks a lot for letting me know. It's great there is a plan for improvement.
As we say in here: ganbarimashō!

Comment 3 clime 2018-04-14 08:40:26 UTC

Hello, actually there was a concrete bug in the current OpenStack setup:

https://pagure.io/fedora-infrastructure/issue/6854

which is now fixed. Situation should be much improved now.

Comment 4 Marc Dequènes (Duck) 2018-04-16 02:20:06 UTC

Thanks a lot :-)