[OpenAFS] Re: Server disk operations speed

Andrew Deason adeason@sinenomine.net
Mon, 8 Apr 2013 13:45:55 -0500


On Mon, 8 Apr 2013 21:09:43 +0300 (EEST)
jukka.tuominen@finndesign.fi wrote:

> The flushed/ not flushed numbers just let me thinking, that - not
> understanding the depths of afs - isn't the afs cache meant to be
> always in sync with the server, or is it possible (default?) for the
> applications to let go the data before it is confirmed to exist on the
> server? That is, if my configuration somehow forces the sync against
> the default, that could explain the poor performance.

When writing to AFS, the unix client will normally store as much of the
'written' data to the local cache as it can before hitting the
fileserver. If the application is issuing fsync() calls, or someone is
closing the file that's being written to, it's possible for us to hit
the server more frequently. That seems unusual for just regular copies,
though.

If you have a very small cache, that would also make you write to the
fileserver more frequently. However, if you're only writing files that
are 150K in size, it's hard to see how that could be an issue. Your
cache size would have to be less than a few MB large for that to be an
impact, I think.

-- 
Andrew Deason
adeason@sinenomine.net