Bug 761944 (GLUSTER-212)

Summary:	replicate read/stat performance deteriorates with increased number of subvolumes
Product:	[Community] GlusterFS	Reporter:	Tomas Zvala <tomas>
Component:	replicate	Assignee:	Vikas Gorur <vikas>
Status:	CLOSED NOTABUG	QA Contact:
Severity:	medium	Docs Contact:
Priority:	low
Version:	2.0.4	CC:	gluster-bugs
Target Milestone:	---
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:		Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:

Description Tomas Zvala 2009-08-13 08:58:39 UTC

Listing all 1500 files in a replicated subvolume scales really bad with increasing number of subvolumes:


4 subvolumes
real	1m9.383s
user	0m0.035s
sys	0m0.121s

3 subvolumes 
real	0m17.322s
user	0m0.028s
sys	0m0.103s

2 subvolumes
real	0m7.287s
user	0m0.024s
sys	0m0.116s

1 subvolume
real	0m2.708s
user	0m0.009s
sys	0m0.104s

native xfs performance
real	0m0.063s
user	0m0.011s
sys	0m0.052s

Comment 1 Vikas Gorur 2009-09-04 04:03:43 UTC

Are these the results of just running 'ls'? Are you sure any self-heal wasn't triggered?

Comment 2 Vikas Gorur 2010-01-28 10:26:16 UTC

We haven't been able to reproduce this in our setup. Closing this bug for now. Please reopen this if you see the issue and you're sure self-heal wasn't
triggered by the 'ls'.

Thanks,
Vikas