[Labs-l] Missing changes in replicas (s2) for 3 hours

[Jaime Crespo]

On the resolution of:

<https://phabricator.wikimedia.org/T129432>

I mention that I have done a "quick fix" that will solve 99% of the
data consistency problems. The rest 1% will be solved in a few weeks,
with a full reimport of all data.

On Sun, Mar 13, 2016 at 6:07 PM, Jaime Crespo <jcrespo at wikimedia.org> wrote:
> For reasons that I am still investigating, database events between:
> 2016-03-08 18:08:24 and 2016-03-08 21:04:38 (all hours UTC) seem to
> not have been replicated to labs-accessible mirrors on s2 wikis [0].
>
> Fix status is tracked on T129432 [1]. I am going to keep providing
> access to those databases because the time frame is relatively small.
> Apologies if your tools return bogus results for operations that
> involves data modified between those times -until the original cause
> is detected and then the issue is corrected.
>
> [0] <https://phabricator.wikimedia.org/diffusion/OMWC/browse/master/dblists/s2.dblist>
> [1] <https://phabricator.wikimedia.org/T129432>
> --
> Jaime Crespo
> <http://wikimedia.org>



-- 
Jaime Crespo
<http://wikimedia.org>