[sheepdog-users] device priority by performance and number of accesses
namei.unix at gmail.com
Thu Jun 6 10:52:00 CEST 2013
On 06/06/2013 12:11 AM, Valerio Pachera wrote:
> I was discussing with Alessandro how to get the best out from some ssd
> we have and wish to use into the cluster.
> Using 250G of disk for cache is even too much and may lead to trouble
> when it's time to flush it.
I am doubtful of the observed 'trouble' for long flushing time. Yeah,
currently if we dd a very big file like 100G, it will flush a long long
time because we currently flush after write finishes. (we can actually
mitigate this problem by introducing background flusher in the object
cache, so it is flushing while you are writing). But for most cases, it
won't cause trouble because most of time, the cache is clean.
More information about the sheepdog-users