Re: Tarsnap deduplication across machines

[Colin Percival <cperciva@tarsnap.com>]

On 01/09/14 10:04, Clayton Davis wrote:
> In place of your:
> 
> (b) copy the cache directories back and forth so that whenever a machine is
> running tarsnap it has an accurate view of the "server state".
> 
> can I assume it's also valid and not-too-expensive to do, in bash, something like:
> 
> alias tarsnap='tarsnap --fsck-prune; tarsnap'
> 
> (Not sure how much bandwidth is used to update a cache directory from the
> tarsnap server.)

That's an option, and it's one which some people on this list have said
in the past that they're using.  Whether it's too expensive, well, that
depends how much you want to spend... the fsck operation needs to download
all the metadata for each archive, which is around 0.1% of the total archive
size.  If you only have a few archives and you're doing this once a week,
probably not a problem; if you have hundreds of archives and you're doing
this every day, it would add up to a lot more.

-- 
Colin Percival
Security Officer Emeritus, FreeBSD | The power to serve
Founder, Tarsnap | www.tarsnap.com | Online backups for the truly paranoid