Please don't reply to lustre-devel. Instead, comment in Bugzilla by using the 
following link:
https://bugzilla.lustre.org/show_bug.cgi?id=11646



(In reply to comment #0)
> I devised a simple test on our test filesystem where:
> 1) multi-stripe file created on client node
> 2) stop lustre on one of the osts where the file resides
> 3) mount filesystem manually
> 4) remove object for file that resides on that ost
> 5) restart lustre
> 6) write to file
> 
> It looks like initially when starting Lustre in this state and we try to write
> to the file, the entire contents of the file hits cache and we see no errors 
> on
> write.  However subsequent writes to the file get -ESTALE.  It seems that the
> memory for this file that hits the cache can now never be reclaimed.  I will
> attach logs and traces with full debug on the issue.
> 
This problem is related to bug 9977 and will be fixed as part of 9977.

_______________________________________________
Lustre-devel mailing list
[email protected]
https://mail.clusterfs.com/mailman/listinfo/lustre-devel

Reply via email to