Bug 762124 (GLUSTER-392)

Summary:	rm -rf fails on distribute setup
Product:	[Community] GlusterFS	Reporter:	Raghavendra G <raghavendra>
Component:	distribute	Assignee:	Anand Avati <aavati>
Status:	CLOSED NOTABUG	QA Contact:
Severity:	low	Docs Contact:
Priority:	low
Version:	mainline	CC:	chrisw, gluster-bugs, krishna
Target Milestone:	---
Target Release:	---
Hardware:	All
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 Krishna Srinivas 2009-11-18 01:53:53 UTC

raghu the reason would be that there would be stale link files in the back end. best way to fix this would be to add selfhealing code to rmdir. before rmdir of dht returns with 'dir not empty' error it can delete the stale link files and try rmdir'ing the backend dirs again.

Comment 1 Raghavendra G 2009-11-18 02:03:30 UTC

root@booradley:/home/raghu/vol-specs/dht# ls /mnt/raghu

root@booradley:/home/raghu/vol-specs/dht# cp -rf /usr/ /mnt/raghu/

root@booradley:/home/raghu/vol-specs/dht# rm -rfv /mnt/raghu/usr  
rm: cannot remove directory `/mnt/raghu/usr': Directory not empty

root@booradley:/home/raghu/vol-specs/dht# ls /mnt/raghu/usr
root@booradley:/home/raghu/vol-specs/dht# 

setup consists of distribute over two clients. ls on backend shows that the directory usr/ is indeed not empty. But readdir is not able to read the dentries properly.

Comment 2 Raghavendra G 2009-11-19 19:28:36 UTC

Did a test again with fresh backend and the bug was not reproducible. Hence closing it.