POL00339056
POL00339056
Message
From: Dave Hulbert [IMCEAEX-
_O=MMS_OU=EXCHANGE+20ADMINISTRATIVE+20GROUP+20+28FYDIBOHF23SPDLT+29_CN=RECIPIENTS_CN=DAVE+2EHULB
ERT8FF12B5A-AD11-4323-AA83-E0874E17B20A@C72A47.ingest.local]
on Dave Hulbert <IMCEAEX-
behalf _O=MMS_OU=EXCHANGE+20ADMINISTRATIVE+20GROUP+20+28FYDIBOHF 23SPDLT+29_CN=RECIPIENTS_CN=DAVE+2EHULB
of ERT8FF12B5A-AD11-4323-AA83-E0874E17B20A@C72A47. ingest.local> [IMCEAEX-
_O=MMS_OU=EXCHANGE+20ADMINISTRATIVE+20GROUP+20+28FYDIBOHF 23SPDLT+29_CN=RECIPIENTS_CN=DAVE+2EHULB
ERT8FF12BSA-AD11-4323-AA83-E0874E17B20A@C72A47.ingest.local]
Sent: 06/05/2012 20:40:17
To: Kevin Gilliland
CC: = Alana Renner
Lesley J Sewell! ‘Antonio Jamasb
+ Neil Lecky-Thompsonj_
Subjec Re: IMPORTANT. Horizon service today during Disaster Recovery test
3
Kevin
sorry for the delayed response. The integrity of data is fully understood by Fujitsu's team and they will
not do anything to compromise it.
A simple explanation of the method being used to correct the data is:
* We use a clean copy of the database in the primary datacentre
+ We apply a fix to correct the indexing issues from today
+ We apply the transactions into the database at the primary site (clean and reindexed).
This way there is no manual intervention and the data remains as per transacted in branch.
Regards
Dave
gement,
Post Office Ltd
- Original Message -----
From: Kevin Gilliland
Sent: Sunday, May 06, 2012 07:21 PM
To: Dave Hulbert
Cc: Alana Renner; Lesley J Sewell; Antonio Jamasb; Neil Lecky-Thompson
Subject: Re: IMPORTANT. Horizon service today during Disaster Recovery test
Is there a risk involved in correcting the transactions before we revert to the primary data centre?
How does this work? Is there any manual intervention? Are we happy that this will have no impact to the
integrity of the system?
Thx,
Kevin
on 6 May 2012, at 13:09, "Dave Hulbert
> Alana, Kevin, Lesley,
>
> The branches have been operating well since they opened this morning; with the Horizon service being
provided out of the secondary data centre.
>
> However, within the data centre, whilst the transactions are completing ok - no issues for branches or
customers - they're not writing to the database correctly. We've been managing this situation but all the
transactions in the database need correcting before we can start the fallback process to the primary data
centre. The volume of transactions has reached the point where the time it will take to correct them will
eat into our fallback time and consequently could impact the service provided tomorrow.
>
> Therefore reluctantly I've agreed to ceasing the service to branches as soon as a smartinform
(telephone) message hits all branches - anticipated to be at approx 13:15. This means we are shortening
the trading day 45 minutes earlier than previously communicated to the branches.
>
POL00339056
POL00339056
> Please accept my apologies for being unable to consult with you all first but an immediate decision was
required.
>
> We will of course provide a full explanation of events later in the week.