[sheepdog-users] Vdi corruption with erasure code if n nodes < x

[Hitoshi Mitake]

On Mon, Oct 6, 2014 at 7:23 AM, Andrew J. Hobbs <ajhobbs at desu.edu> wrote:
> Would it be possible to freeze erasure encoded vdis rather than halting
> the cluster outright?

I looked at the recovery code and correct recovery process for erasure
coding is already implemented. It doesn't works because of some bugs.
I'll fix it so halting cluster or freezing VDIs wouldn't be required.
# I'm on a business trip to U.S. so my reply and development is delaying, sorry

Thanks,
Hitoshi

>
> On 10/05/2014 02:33 AM, Hitoshi Mitake wrote:
>> On Wed, Sep 10, 2014 at 4:30 PM, Valerio Pachera <sirio81 at gmail.com> wrote:
>>>
>>> 2014-09-10 9:30 GMT+02:00 Valerio Pachera <sirio81 at gmail.com>:
>>>> I opend the bug in launch pad.
>>>> Please, set "Critical" importance to it.
>>>
>>> Forgot to paste :-)
>>> https://bugs.launchpad.net/sheepdog-project/+bug/1367612
>> I looked at this problem and it is tightly related to current recovery
>> scheme. Solving this problem requires amount of change for recovery.
>>
>> Slack way to avoid the problem is shutting down your cluster when a
>> number of dead node becomes y (y is the smallest x:y on your cluster).
>>
>> I think the above slack solution can be implemented in 0.9. It would
>> take one week so I have to delay v0.9.0. How do you think?
>>
>> Thanks,
>> Hitoshi
>>
>>>
>>> --
>>> sheepdog-users mailing lists
>>> sheepdog-users at lists.wpkg.org
>>> http://lists.wpkg.org/mailman/listinfo/sheepdog-users
>>>
>
>
> --
> sheepdog-users mailing lists
> sheepdog-users at lists.wpkg.org
> http://lists.wpkg.org/mailman/listinfo/sheepdog-users
>