Hi Greg,
On Mon, Apr 04, 2005 at 01:45:41PM -0400, Gregory J. Sharp wrote:
> I tried the 64K release of xrootd. The obvious stuff all seems to work.
> (Took me a little while to spot that I needed a -t flag on the xrootd
> data servers, but once I woke up to that it ran fine.) The performance
> seems better than before, although my new hardware setup was likely to
> produce better performance. (We "optimized" the network between nodes
> to put them all on the same GB Ethernet switch card).
I'm not sure which was the last release you tested, but since they are
tuning up for the "big memory machine" at SLAC, Andy has done a lot
of work in recent months for I/O optimization and reducing the memory
footprint. (I think the requirement for the "-t" option was also added
since early last fall. Derek Feichtinger pointed out a couple of weeks
ago that the "multiple servers" example no longer worked because this was
missing. I just finally updated that today.)
> I will acquaint myself with the code and see what I can find :-)
Excellent, thanks... ;-)
Pete
-------------------------------------------------------------------------
Peter Elmer E-mail: [log in to unmask] Phone: +41 (22) 767-4644
Address: CERN Division PPE, Bat. 32 2C-14, CH-1211 Geneva 23, Switzerland
-------------------------------------------------------------------------
