Bug 1285289 - Files skipped during rebalance because destination brick size is lesser than source brick size
Files skipped during rebalance because destination brick size is lesser than ...
Status: CLOSED NOTABUG
Product: Red Hat Gluster Storage
Classification: Red Hat
Component: distribute (Show other bugs)
3.1
Unspecified Unspecified
unspecified Severity medium
: ---
: ---
Assigned To: Nithya Balachandran
storage-qa-internal@redhat.com
: Triaged
Depends On:
Blocks:
  Show dependency treegraph
 
Reported: 2015-11-25 06:17 EST by Shruti Sampat
Modified: 2016-06-28 03:15 EDT (History)
2 users (show)

See Also:
Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of:
Environment:
Last Closed: 2016-06-28 03:15:13 EDT
Type: Bug
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:
Category: ---
oVirt Team: ---
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---


Attachments (Terms of Use)

  None (edit)
Description Shruti Sampat 2015-11-25 06:17:36 EST
Description of problem:
-----------------------

While running rebalance on a 3x2 distributed replicate volume, I see that files being skipped. The following is from rebalance logs -

[2015-11-25 16:29:59.914769] I [dht-rebalance.c:1002:dht_migrate_file] 0-vol-dht: /newfile_944: attempting to move from vol-replicate-1 to vol-replicate-0
[2015-11-25 16:29:59.914796] I [dht-rebalance.c:1002:dht_migrate_file] 0-vol-dht: /newfile_952: attempting to move from vol-replicate-1 to vol-replicate-0
[2015-11-25 16:29:59.921714] W [MSGID: 109023] [dht-rebalance.c:657:__dht_check_free_space] 0-vol-dht: data movement attempted from node (vol-replicate-1:142365920) with higher disk space to a node (vol-repli
cate-2:141751520) with lesser disk space, file { blocks:204800, name:(/newfile_927) }
[2015-11-25 16:29:59.922552] I [dht-rebalance.c:1002:dht_migrate_file] 0-vol-dht: /newfile_967: attempting to move from vol-replicate-1 to vol-replicate-2
[2015-11-25 16:29:59.931073] W [MSGID: 109023] [dht-rebalance.c:657:__dht_check_free_space] 0-vol-dht: data movement attempted from node (vol-replicate-1:142365920) with higher disk space to a node (vol-replicate-0:142161120) with lesser disk space, file { blocks:204800, name:(/newfile_952) }
[2015-11-25 16:29:59.931814] W [MSGID: 109023] [dht-rebalance.c:657:__dht_check_free_space] 0-vol-dht: data movement attempted from node (vol-replicate-1:142365920) with higher disk space to a node (vol-replicate-0:142161120) with lesser disk space, file { blocks:204800, name:(/newfile_944) }
[2015-11-25 16:29:59.939816] W [MSGID: 109023] [dht-rebalance.c:657:__dht_check_free_space] 0-vol-dht: data movement attempted from node (vol-replicate-1:142365920) with higher disk space to a node (vol-replicate-2:141751520) with lesser disk space, file { blocks:204800, name:(/newfile_967) }

I see these files to be present as `T' files on the destination brick.

The destination brick has enough space to accommodate the file so it should be migrated, in my opinion. 

Version-Release number of selected component (if applicable):
--------------------------------------------------------------
glusterfs-3.7.1-11.el7rhgs.x86_64

How reproducible:
-----------------
Frequently

Steps to Reproduce:
-------------------
1. A 2x2 distributed-replicate volume running inside a container is mounted on another container.  
2. Created about 1000 files of 100M each. 
3. Added another pair of bricks to make the volume 3x2.
4. Performed rebalance on the volume.
5. After the rebalance was completed, renamed each file from the client container.
6. Started rebalance on the volume again.

Actual results:
---------------
Some files were skipped during migration.

Expected results:
-----------------
Files should not be skipped as long as the destination brick has sufficient space to accommodate it.

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