Unfortunately, the tool the dynamically enable/disable image features (rbd 
feature disable <image-spec> <feature>) was added during the Infernalis 
development cycle.  Therefore, in the short-term you would need to recreate the 
images via export/import or clone/flatten.  

There are several object map / exclusive lock bug fixes that are scheduled to 
be included in the 0.94.4 release that might address your issue.  However, 
without more information, we won't know if the issue you are seeing has been 
fixed or not in a later release.

If it is at all possible, it would be most helpful if you could provide logs or 
a backtrace when the VMs lock up.  Since it happens so infrequently, you may 
not be willing to increase the RBD debug log level to 20 on one of these VMs. 
Therefore, another possibility is for you to run "gcore <VM qemu pid>" to 
generate a core dump or attach GDB to the hung VM and run "thread apply all 
bt".  With the gcore or backtrace method, we would need a listing of all the 
package versions installed on the machine to recreate a similar debug 
environment.

Thanks,

Jason 


----- Original Message -----
> From: "Christoph Adomeit" <christoph.adom...@gatworks.de>
> To: ceph-users@lists.ceph.com
> Sent: Monday, August 31, 2015 7:49:00 AM
> Subject: [ceph-users] How to disable object-map and exclusive features ?
> 
> Hi there,
> 
> I have a ceph-cluster (0.94-2) with >100 rbd kvm images.
> 
> Most vms are running rock-solid but 7 vms are hanging about once a week.
> 
> I found out the hanging machines have
> features: layering, exclusive, object map while all other vms do not have
> exclusive and object map set.
> 
> Now I want to disable these features. Is ist possible to disable these
> features while the vms are running ? Or at least while they are shut down ?
> Or will I have to recreate all these images ?
> 
> Thanks
>   Christoph
> _______________________________________________
> ceph-users mailing list
> ceph-users@lists.ceph.com
> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
> 
_______________________________________________
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

Reply via email to