[Gluster-devel] would an io-disk-cache be feasible?

[Anand Avati]

Sascha,
 this idea has come up many times in our internal discussions. it surely is
a feasible idea, but it just has not earned enough priority yet to get into
our roadmap.

avati

2008/1/9, Sascha Ottolski <ottolski at web.de>:
>
> Hi,
>
> I'm wondering how complicated it woul be to create a translator that
> would cache data on the local disk. Especially in our use case I think
> this would reduce the gluster and network load immense. But I might be
> wrong.
>
> What is special in our setup is, that the files we put on the cluster do
> never change; some get deleted somtimes, though, but seldom. Most of
> them stay forever, writes are low, reads are high. I still have no good
> idea about how often a "cache hit" might occur, but surely some of the
> files will have them. The files itself are rather small (images of some
> kb or small text files of only a few 10 bytes).
>
>
> Thanks, Sascha
>
>
>
> _______________________________________________
> Gluster-devel mailing list
> Gluster-devel at nongnu.org
> http://lists.nongnu.org/mailman/listinfo/gluster-devel
>



-- 
If I traveled to the end of the rainbow
As Dame Fortune did intend,
Murphy would be there to tell me
The pot's at the other end.