On Sun, Dec 17, 2006 at 03:36:43PM -0500, Scott Atchley wrote:
> What do I need to do to include bmi_mx in the configure and make  
> process? Would it be easier to work with a cvs checkout or a release  
> tarball?

Definitely work from a CVS checkout if you can.  I don't think there
have been many bmi-specific changes, but it will make the eventual
merge much easier.

> As for testing, I do not have disks fast enough to stress  
> Myrinet-2000 (2Gb/s) NICs let alone Myri-10G NICs. In other emails,  
> these options have been mentioned. Which makes most sense (least  
> amount of coding, compiling issues, etc.)? Do they require disk IO?

The methods you outlined below (flowproto-dump-offsets, the tests in
test/io/bmi, the TAS trove implementation) will take a bit of work.

One approach that works well for people is to set up the pvfs storage
space in a tmpfs or ramdisk.  Then you're bounded only by memory
bandwidth, and you'll be working with the same code paths that you'd
use in production.

==rob

-- 
Rob Latham
Mathematics and Computer Science Division    A215 0178 EA2D B059 8CDF
Argonne National Lab, IL USA                 B29D F333 664A 4280 315B
_______________________________________________
Pvfs2-developers mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers

Reply via email to