> -----Original Message-----
> From: GRIDPP2: Deployment and support of SRM and local storage
> management [mailto:[log in to unmask]] On Behalf Of John
>
> Sorry I wasn't at the meeting, missed the email. I noticed in the
> minutes the subject of moving from rfcp to RFIO with 4kB buffers.
We've
> been looking into the problems of rfcp at Liverpool and have come to
> some conclusions about the lack of performance and ways of improving
> efficiency.
>
I'm going to take some time to think this over properly, but I have a
couple
of quick questions:
> Following system activity during analysis shows that while there is a
> lot of read access to the WN disk the major disk activity (at least
> twice as much and often more) is *write* access. See this plot for a
WN
> disk
>
> http://hep.ph.liv.ac.uk/~jbland/device_i_o_transfers__dev9_0.png
- That's a nice plot, how did you generate it?
- Was this during one of the current round of rfcp file stager based
tests? If it was then I wonder how much of the write traffic is the
input files being staged in, as opposed to actual results being written
out? Were we to move back to rfio posix access I'd expect the former
to be eliminated/pushed out to the disk pools.
Ewan
|