Bug 1181588 - RFE: Rebalance shouldn't start when clients older than glusterfs-3.6 are connected
Summary: RFE: Rebalance shouldn't start when clients older than glusterfs-3.6 are conn...
Keywords:
Status: CLOSED WONTFIX
Alias: None
Product: GlusterFS
Classification: Community
Component: glusterd
Version: 3.6.1
Hardware: Unspecified
OS: Unspecified
high
medium
Target Milestone: ---
Assignee: bugs@gluster.org
QA Contact:
URL:
Whiteboard:
Depends On: 1136201
Blocks: glusterfs-3.6.2
TreeView+ depends on / blocked
 
Reported: 2015-01-13 12:28 UTC by Niels de Vos
Modified: 2016-08-01 04:43 UTC (History)
5 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of: 1136201
Environment:
Last Closed: 2016-08-01 04:42:13 UTC
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:
Embargoed:


Attachments (Terms of Use)

Description Niels de Vos 2015-01-13 12:28:20 UTC
+++ This bug was initially created as a clone of Bug #1136201 +++
+++                                                           +++
+++ Please use this bug to backport related fixes to 3.6      +++
+++                                                           +++

The DHT xlator has been fixed in glusterfs-3.6 and several data race issues have been fixed which used to cause data loss during rebalance. With these changes, glusterfs-3.6 and newer clients can correctly co-operate with the rebalance process and not cause any data loss. But clients older than glusterfs-3.6 still have the bugs and could cause data loss when rebalance happens.

To prevent the above, rebalance on glusterfs-3.6 and above need to start only if older clients haven't mounted a volume. The gluster cli needs to warn the user of older clients and fail when attempting a rebalance start. Also, the user should be able to bypass the check using 'force' if needed.

--- Additional comment from Anand Avati on 2014-09-03 07:55:03 CEST ---

REVIEW: http://review.gluster.org/8583 (glusterd: Prevent rebalance starting with old clients) posted (#1) for review on master by Kaushal M (kaushal)

--- Additional comment from Anand Avati on 2014-09-03 11:38:12 CEST ---

COMMIT: http://review.gluster.org/8583 committed in master by Krishnan Parthasarathi (kparthas) 
------
commit 312b339190b7b8f2e38cdd210bf2ee71e53b287e
Author: Kaushal M <kaushal>
Date:   Wed Sep 3 10:50:31 2014 +0530

    glusterd: Prevent rebalance starting with old clients
    
    Glusterd will prevent rebalance from starting when clients older than
    glusterfs-v3.6.0 are connected to a volume. This is needed as running
    rebalance with old clients connected could lead to data loss in some
    cases. The DHT xlator on newer clients (>= 3.6.0) has been fixed to
    prevent the data loss issues.
    
    Change-Id: If58640236382a2fc13f73f6b43777f01713859f7
    BUG: 1136201
    Signed-off-by: Kaushal M <kaushal>
    Reviewed-on: http://review.gluster.org/8583
    Tested-by: Gluster Build System <jenkins.com>
    Reviewed-by: Atin Mukherjee <amukherj>
    Reviewed-by: Krishnan Parthasarathi <kparthas>
    Tested-by: Krishnan Parthasarathi <kparthas>

--- Additional comment from Atin Mukherjee on 2014-09-04 11:45:19 CEST ---

Patch has made into codebase.

Comment 1 Atin Mukherjee 2016-08-01 04:42:13 UTC
This is not a security bug, not going to fix this in 3.6.x because of
http://www.gluster.org/pipermail/gluster-users/2016-July/027682.html

Comment 2 Atin Mukherjee 2016-08-01 04:43:46 UTC
If the issue persists in the latest releases, please feel free to clone them


Note You need to log in before you can comment on or make changes to this bug.