[OpenAFS-devel] STABLE14-background-fsync-consistency-issues-20090522 - can this be?

[Rainer Toebbicke]

Yes, silly me I got confused hunting down a strange problem in that area which 
made me forget that volume special files are closed but the handles kept 
around on purpose.

That problem, by they way, manifests itself as volume going offline, salvager 
does not find anything, but volume going offline again. This time, the problem 
is not a wrong magic in the link table which volinfo -fixheader would correct. 
   Usually, the users see messages like "unable to remove file - disk space 
exceeded" - certainly a message to get confused.

It goes away if you move the volume to another server. i.e. the volume is ok. 
It re-appears if you move the volume back to the *same* partition on the same 
server. We found in the dumps that after the move ihandles for the volume 
header stay around - this smells like a leak which made me jump on that patch 
too quickly.

I'll engage brain before touching keyboard next time - sorry for that.




Jeffrey Altman schrieb:
> Rainer Toebbicke wrote:

> 
> I believe the patch is correct.   IH_REALLYCLOSE() will close the file
> descriptors but not release the inode handle.  Wherease IH_RELEASE()
> will close the file descriptors and release the inode handle.
> 
> Likewise, VCloseVolumeHandles_r() is closing the file descriptors but
> intentionally leaving the inode handles alone.
> VReleaseVolumeHandles_r() is the equivalent function that is used to
> release the inode handles.
> 


-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Rainer Toebbicke
European Laboratory for Particle Physics(CERN) - Geneva, Switzerland
Phone: +41 22 767 8985       Fax: +41 22 767 7155