Bug 1146256

Summary:	Initial Georeplication fails to use correct GID on folders ONLY
Product:	[Red Hat Storage] Red Hat Gluster Storage	Reporter:	Harold Miller <hamiller>
Component:	geo-replication	Assignee:	Aravinda VK <avishwan>
Status:	CLOSED ERRATA	QA Contact:	storage-qa-internal <storage-qa-internal>
Severity:	high	Docs Contact:
Priority:	high
Version:	2.1	CC:	aavati, avishwan, csaba, fharshav, nlevinki, psriniva, smanjara, surs, vagarwal
Target Milestone:	---	Keywords:	ZStream
Target Release:	RHGS 2.1.6
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:	glusterfs-3.4.0.71rhs-1.el6rhs	Doc Type:	Bug Fix
Doc Text:	Previously, Geo-replication sessions failed to synchronize the ownerships of empty files or files copied from other location. The files in the slave volume had different ownerships and permissions. With this fix, files in both master and slave volumes have the same ownership and permission.	Story Points:	---
Clone Of:
Clones:	1146263 (view as bug list)		Environment:
Last Closed:	2015-01-28 10:42:27 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:	1104061
Bug Blocks:	1146263, 1176659

Description Harold Miller 2014-09-24 21:00:04 UTC

Description of problem: Fresh geo-replication fails to copy directory group to slave.


Version-Release number of selected component (if applicable): 3.4.0.61geo-1.el6rhs.1.hotfix.sfdc01121015.sfdc0117419


Steps to Reproduce: (Custoemrs words)

1.We deleted the geo-replication sessions, destroyed the slave volumes, reformatted the brick file systems, rebuilt the slave volumes, and recreated geo-replication sessions to sync to the completely empty slave volumes.

2.After the hybrid crawl finished, we used a read-only rsync command to compare the master and slave. 

Actual results:We found that a large proportion of directories copied to the slave volume did not have the correct group. The GID of the directory in the slave volume was instead set to the UID of the directory owner.

*Files* did not exhibit this behavior, just directories.

Expected results: UID and GID on files and folders should remain unchanged


Additional info:

Comment 3 Aravinda VK 2014-12-01 05:35:48 UTC

Patches sent to downstream:
https://code.engineering.redhat.com/gerrit/#/c/37445/
https://code.engineering.redhat.com/gerrit/#/c/37540/

Comment 5 shilpa 2015-01-09 07:28:45 UTC

Verified on glusterfs-3.4.0.71

# cd /mnt/master
# mkdir deep1
# useradd shilpa
# chown -R shilpa deep1
# su shilpa
# mkdir deep2

On master:
# ls -l
drwxr-xr-x 2 shilpa root 12 Jan  9 07:05 deep1

$ ls -l /mnt/master/level0/
total 0
drwxrwxr-x 3 shilpa shilpa 38 Jan  9 07:14 level1

On slave:
# ls -l /mnt/slave
drwxr-xr-x 3 shilpa root 36 Jan  9 07:08 deep1

$ ls -l /mnt/slave/level0/
total 0
drwxrwxr-x 3 shilpa shilpa 38 Jan  9 07:14 level1

Comment 7 Pavithra 2015-01-16 11:10:05 UTC

Hi Aravinda,

I removed some internal details of how geo-rep works. I've edited the doc text. Can you please review it and sign off?

Comment 10 errata-xmlrpc 2015-01-28 10:42:27 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/RHBA-2015-0095.html