[Gluster-devel] Re: Strange behavior of AFR

[Anton Khalikov]

> This may be to do with tthe posix locking. Currently posix lock server is 
> the first server in the AFR list (and the order of servers in AFR whould 
> be the same on all nodes, or else the locking won't work properly). When 
> the primary (lock) server goes away, all the locks disappear, too. There 
> was also another thread there discussing lock / metadata distribution 
> across the cluster with quorum locking. But that is also as yet 
> unimplemented.

Do you mean posix-locks translator or what ? I don't use it by the way.



> If it's timeout related, then yes, head-ing the images will solve the 
> problem. If it's lock related, it won't make any difference.

Yes it is timeout related. XEN tools as well as domU kernels try to
reach filesystem and their requests are blocked until resync is 100%
complete. It raises timeouts.


> For the particular use case you are describing, you may find that 
> something like DRBD would fit better, with a separate DRBD device per 
> VM.

I thought there are success stories with glusterfs and xen ?
If not, I'll try to look at DRBD. Thanks for your suggestion.

--
Best regards
Anton Khalikov