Bug 762054 (GLUSTER-322)

Summary:	Race condition among operations waiting for completion of open in a setup having replicate below quick-read.
Product:	[Community] GlusterFS	Reporter:	Raghavendra G <raghavendra>
Component:	quick-read	Assignee:	Raghavendra G <raghavendra>
Status:	CLOSED NOTABUG	QA Contact:
Severity:	low	Docs Contact:
Priority:	low
Version:	mainline
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 Raghavendra G 2009-10-15 10:43:18 UTC

Marking this bug as invalid, since there will be no more than one operation from a single thread in the waiting list of operations waiting for open. As for as the order of operations between two different goes, quick read need not worry about their order, since the threads themselves have to synchronize the order.

Comment 1 Raghavendra G 2009-10-15 13:35:33 UTC

The operations waiting for completion of open are stack_wound in a loop in qr_open_cbk. With afr below quick-read, there can be race-conditions among the operations since in afr, an fop is implemented as a transaction.

The fix should resume the operations in waiting list one by one making sure that the reply to the first operation in list is got before resuming the next operation.