[sheepdog-users] [sheepdog] Dead Sheeps...

[Liu Yuan]

On 06/19/2012 04:58 PM, Bastian Scholz wrote:
> after running the tests and all virtual machines successfully with
> the temporary fix in devel branch for one week, I updated sheepdog
> to actual master branch and the error (at a first view the same)
> reappeared.
> 
> Before start debugging I would ask, if the temporary fix is already
> in master branch (I guess yes).

Not yet, I just removed the offending code which tries to solve a
partial write to all replica because of node failure. But I don't come
up with a better approach to solve the partial write problem (which
actually won't cause trouble mostly because the VM will receive a EIO
for this case). So I need Kazutaka's option about removal of
fix_object_consistency() solely.

Thanks,
Yuan