On Jan 5, 2007, at 11:38 AM, Scott Atchley wrote:
On Dec 29, 2006, at 11:44 AM, Pete Wyckoff wrote:
pvfs2-cp isn't that great a code. Find yourself an MPI interface
benchmark, like "perf". This produces server load around 90%:
ib30$ mpiexec -n 1 2402/perf -n 10 -s 800m -c 100m -f pvfs2:/pvfs-
ib/x1
#np size chunk write no sync- read no sync-- write sync----
read sync-----
# (MB) (MB) (MB/s) (MB/s) (MB/s) (MB/s)
1 800.0 100.0 681.56 +- 1.9 612.87 +- 2.2 679.99 +- 1.0
613.76 +- 3.0
With 1 MB flow buffers, the server is pegged and slower, more like
what you're seeing:
ib30$ mpiexec -n 1 2402/perf -n 10 -s 800m -c 100m -f pvfs2:/pvfs-
ib/x1
#np size chunk write no sync- read no sync-- write sync----
read sync-----
# (MB) (MB) (MB/s) (MB/s) (MB/s) (MB/s)
1 800.0 100.0 342.73 +- 3.8 317.24 +- 2.6 343.96 +- 2.2
318.34 +- 1.8
I modified perf.c to write 10 times a buffer of 128 MBs. I get:
% mpiexec -n 1 ./perf -fname pvfs2:/mnt/pvfs2/x1
Access size per process = 134217728 bytes, ntimes = 10
Write bandwidth without file sync = 827.824860 Mbytes/sec
Read bandwidth without prior file sync = 870.979531 Mbytes/sec
Write bandwidth including file sync = 827.166730 Mbytes/sec
Read bandwidth after file sync = 865.624838 Mbytes/sec
using a single client. The client used about 25% CPU.
Scott
_______________________________________________
Pvfs2-developers mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers
