[PATCH 00/14] repair memory usage reductions
Christoph Hellwig
hch at infradead.org
Fri Sep 4 09:51:09 CDT 2009
On Fri, Sep 04, 2009 at 08:37:37AM -0500, Geoffrey Wehrman wrote:
> On Fri, Sep 04, 2009 at 12:57:53PM +1000, Dave Chinner wrote:
> | Christoph asked me to repeat what I said on #xfs w.r.t the regression.
>
> Thank you for the detailed description. All I had was a statement from
> January 2008, "Barry has completed the memory optimization, but initial
> testing shows that performance has regressed." That was the last update
> recorded on Barry's work.
>
> | With that in mind, I think the memory usage optimisation is far more
> | important to the majority of XFS users than the CPU usage regression
> | it causes as the majority of users don't have RAM-rich environments
> | to run repair in.
>
> I agree.
In my testing I haven't seen big differences in performance, it
sometimes got a bit faster and sometimes a bit slower. I will send out
a more detailed performace report in a few days.
More information about the xfs
mailing list