Fergus Henderson wrote:
The include server could be the bottleneck. What's the CPU usage for
the include server process?
Or it could be disk I/O. Try iostat or vmstat to profile that.
After some more experimentation, I think I may have a clue what's going
on here. I think I may be getting bound up in context switching costs,
both in my single node performance and on localhost using distcc.
On a single node:
-j4: 24 million context switches @ 10us = 4m
-j8: 96 million context switches @ 10us = 16m
Command being timed: "make -j4 debug"
User time (seconds): 378.81
System time (seconds): 566.57
Percent of CPU this job got: 265%
Elapsed (wall clock) time (h:mm:ss or m:ss): 5:55.91
Major (requiring I/O) page faults: 472
Minor (reclaiming a frame) page faults: 26761810
Voluntary context switches: 24629071
Involuntary context switches: 269418
Command being timed: "make -j8 debug"
User time (seconds): 427.55
System time (seconds): 2250.72
Percent of CPU this job got: 429%
Elapsed (wall clock) time (h:mm:ss or m:ss): 10:23.29
Major (requiring I/O) page faults: 270
Minor (reclaiming a frame) page faults: 26760678
Voluntary context switches: 95391888
Involuntary context switches: 825155
I don't understand why doubling the job count increases context
switching by 4x. Any insights appreciated.
Thanks,
--
Robert W. Anderson
Center for Applied Scientific Computing
Email: anderson...@llnl.gov
Tel: 925-424-2858 Fax: 925-423-8704
__
distcc mailing list http://distcc.samba.org/
To unsubscribe or change options:
https://lists.samba.org/mailman/listinfo/distcc
