Bug 1229567

Summary:	context of access control translator should be updated properly for GF_POSIX_ACL_*_KEY xattrs
Product:	[Red Hat Storage] Red Hat Gluster Storage	Reporter:	Jiffin <jthottan>
Component:	glusterfs	Assignee:	Jiffin <jthottan>
Status:	CLOSED ERRATA	QA Contact:	Apeksha <akhakhar>
Severity:	high	Docs Contact:
Priority:	high
Version:	rhgs-3.1	CC:	akhakhar, annair, ansubram, asrivast, jthottan, kkeithle, mmadhusu, ndevos, nlevinki, saujain, skoduri, vagarwal, vbellur
Target Milestone:	---
Target Release:	RHGS 3.1.0
Hardware:	All
OS:	All
Whiteboard:
Fixed In Version:	glusterfs-3.7.1-8	Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:	1226807	Environment:
Last Closed:	2015-07-29 04:59:41 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:
Bug Depends On:	1215174, 1226807, 1228155, 1229564, 1229860, 1230327
Bug Blocks:	1202842

Description Jiffin 2015-06-09 06:02:06 UTC

+++ This bug was initially created as a clone of Bug #1226807 +++

Description of problem:
Ganesha caches the ACLs for a directory too long, needs some re-validation or something. nfs4_setfacl will fails due to that. but succeeds when nfs-ganesha is restarted or when there is a sufficient delay.

Version-Release number of selected component (if applicable):
mainline

How reproducible:
always

Steps to Reproduce:
1.Create and start a volume

2.Export a volume using nfs-ganesha

3.Create any user(user1) and group(group1) in the server and client

4.Mount the volume using nfs4 protocol

5.create directory at mount point : mkdir src

6.set acl(default acl for directory) using nfs4_setfacl on src: nfs4_setfacl -a A:gdf:group1:RWX src 

Actual results:

nfs4_setfacl fails, but succeeds when there is a sufficient delay or when ganesha is restarted.

Expected results:

nfs4_setfacl should succeed

Additional info:
The POSIX ACLs on the brick are set correctly. May be need to invalidate the dentry.

Comment 4 Jiffin 2015-06-16 06:10:58 UTC

This is a known issue , should be fixed in next release.

Comment 6 Jiffin 2015-06-24 13:20:45 UTC

It is related to access control translator in gluster code, when an acl is set on directory , it update the in memory context of access control translator.
A simple workaround for this issue is perform 'ls' on the directory after setting the acl.

Comment 10 Jiffin 2015-07-04 09:01:11 UTC

The patch is posted at https://code.engineering.redhat.com/gerrit/#/c/52284/

Comment 11 Jiffin 2015-07-16 08:36:40 UTC

This one fixed is in https://code.engineering.redhat.com/gerrit/#/c/52284/

Comment 12 Apeksha 2015-07-20 07:05:54 UTC

nfs4_setfacl succeeds.
    
Output:

    # mkdir /mnt/vol5/src1
     
    # nfs4_setfacl -a A:gdf:acl_group.blr.redhat.com:RWX /mnt/vol5/src1
    # nfs4_getfacl /mnt/vol5/src1/
    A::OWNER@:rwaDxtTcCy
    A::GROUP@:rxtcy
    A:g:acl_group.blr.redhat.com:rwaDxtcy
    A::EVERYONE@:rxtcy
    A:fdi:OWNER@:tcy
    A:fdi:GROUP@:tcy
    A:fdig:acl_group.blr.redhat.com:rwaDxtcy
    A:fdi:EVERYONE@:tcy

Comment 13 errata-xmlrpc 2015-07-29 04:59:41 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://rhn.redhat.com/errata/RHSA-2015-1495.html