[LU-161] connectathon test on NFS v3 client failed Created: 24/Mar/11  Updated: 30/Mar/11  Resolved: 27/Mar/11

Status: Resolved
Project: Lustre
Component/s: None
Affects Version/s: Lustre 2.1.0
Fix Version/s: Lustre 2.1.0

Type: Bug Priority: Blocker
Reporter: Sarah Liu Assignee: nasf (Inactive)
Resolution: Fixed Votes: 0
Labels: None

Severity: 3
Rank (Obsolete): 5188

 Description   

connectathon special test failed on two NFS v3 clients.

test log
------------------
test truncate
truncate succeeded

test holey file support
read (hole) offset 12513, sz = 56667, bytes = 820 (ret -1), holesz = 9012
read: Not a directory
special tests failed
parallel-scale test_connectathon: @@@@@@ FAIL: connectathon failed: 1



 Comments   
Comment by Di Wang [ 24/Mar/11 ]

The reason is that we enable ATTR_FID in ORNL-2, but did not handle getattr_lock req well with duplicate fids in RPC, (please refer how 1.8 handle this if there are duplicate fids in getattr_lock rpc). Fang yong, could you please look at this? I would guess that is also the reason of LU-142, LU-152. Thanks.

Comment by nasf (Inactive) [ 24/Mar/11 ]

The patch is in inspection:

http://review.whamcloud.com/#change,365

Comment by Sarah Liu [ 25/Mar/11 ]

this bug and LU-142 are fixed with server build: http://build.whamcloud.com/job/reviews-centos5/562/ and client build: http://build.whamcloud.com/job/reviews-patchless-centos5/282/

Comment by Peter Jones [ 26/Mar/11 ]

Reopen ticket until patch has landed

Comment by nasf (Inactive) [ 27/Mar/11 ]

The patch is landed.

Comment by Sarah Liu [ 30/Mar/11 ]

verified.
http://build.whamcloud.com/view/Lustre%202.x/job/lustre-master-centos5/171/
http://build.whamcloud.com/view/Lustre%202.x/job/lustre-master-client-centos5/133/

Generated at Sat Feb 10 01:04:21 UTC 2024 using Jira 9.4.14#940014-sha1:734e6822bbf0d45eff9af51f82432957f73aa32c.