Bug 216168

Summary: ACL on NFS failed intermittetly
Product: Red Hat Enterprise Linux 4 Reporter: Shing-Shong Shei <shei>
Component: kernelAssignee: Red Hat Kernel Manager <kernel-mgr>
Status: CLOSED NEXTRELEASE QA Contact: Red Hat Kernel QE team <kernel-qe>
Severity: medium Docs Contact:
Priority: low    
Version: 4.4CC: kdudka, sct
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2009-12-17 15:22:40 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
The strace outupt when it failed.
none
strace output when it works none

Description Shing-Shong Shei 2006-11-17 16:31:42 UTC 
Description of problem: ACL works intermittently via NFS


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


How reproducible: Not sure -- don't know how it failed and how it started
working again.  Two scripts enclosed.


Steps to Reproduce:
1.
2.
3.
  
Actual results:


Expected results:


Additional info:

We have a RHEL 4.4 NFS file server and when users try to use ACL on the
client machines, sometimes it works but sometimes it failed.  We don't
have a way to reproduce it reliably yet.  Luckily we caught two instances
using strace that shows a successful one and a failed one.
Comment 1 Shing-Shong Shei 2006-11-17 16:31:42 UTC 
Created attachment 141492 [details]
The strace outupt when it failed.
Comment 2 Shing-Shong Shei 2006-11-17 16:32:35 UTC 
Created attachment 141493 [details]
strace output when it works
Comment 3 Shing-Shong Shei 2006-12-02 16:11:17 UTC 
Anybody home?  Has this case been taken?
Thanks.  --Bruce
Comment 4 Kamil Dudka 2009-12-17 15:01:18 UTC 
In both cases strace output is exactly the same until the first getxattr syscall, which returns ENODATA. This means it can't be a bug of getfacl. It might be caused by NFS or more likely the remote server.

Any chance the remote file could be alternated between the tests?
Comment 5 Shing-Shong Shei 2009-12-17 15:11:16 UTC 
> Any chance the remote file could be alternated between the tests?

Wow three years later, finally somebody starts to look into this issue.  Frankly speaking, it's been so long that I cannot even recall the details.  But if you want an answer, I'd say that the remote file has not been changed for sure as during the test, no reason for me to change it, especially when I tried to make a case.  Anyway, we have since moved to RHEL 5 and have not seen this error.  So we can close this case now.  Thanks for looking into it.  Have a Happy Holidays.
Comment 6 Kamil Dudka 2009-12-17 15:22:40 UTC 
Sorry for the delay. Don't blame me as I got the ACL package three hours ago :-) Thanks for the additional info. I am closing the bug as NEXTRELEASE.