Re: Interrupted backups

[John Gamble <jg5@sanger.ac.uk>]

Hi Arnt,

Thanks for your reply.  The example I quoted assumed that Wednesday was the first ever backup.  I should have made that clearer, sorry.

Regards,

John

##

On 20 Mar 2014, at 14:30, Arnt Gulbrandsen <arnt@gulbrandsen.priv.no> wrote:

That's not entirely how it works.

On Wednesday, the tarsnap client needed to make a backup and started computing the set of blocks needed for that. Some of those blocks had already been uploaded, others it uploaded.

On Thursday, the tarsnap client needed to make a backup and started computing the set of blocks needed for that. Some of those blocks had already been uploaded, others it uploaded. When it was done, it recorded the fact that Thursday's backup was complete.

Thursday's tarsnap invocation could not tell whether the already-uploaded blocks were uploaded on Wednesday or earlier. It just knew that the blocks were needed for Thursday's backup, and were present on the server since one or more earlier backups needed them.

(I'm lying very slightly.)

Arnt

--------------------

John Gamble

Senior Computer Biologist
Cancer Genome Project
Wellcome Trust Sanger Institute
Cambridge,  UK
CB10 1SA

Tel: +44 (0)1223 - 834244
Ext: 7703
jg5@sanger.ac.uk

-- The Wellcome Trust Sanger Institute is operated by Genome Research Limited, a charity registered in England with number 1021457 and a company registered in England with number 2742969, whose registered office is 215 Euston Road, London, NW1 2BE.