> 2.3.11 had been up and running for just over 48 hours, during which I
> noticed that the machine performance was slowly but surely declining.
> New processes were taking longer to launch, and existing processes were
> running *significantly* slower. In particular, "rc5des" normally
> processes around 335 kkeys/sec on this machine, and performance had
> dropped to around 107 kkeys/sec. The "top" command showed "update"
> eating most of the available resources (followed by "rc5des"), whereas
> in prior kernel releases, "update" doesn't even make it onto the list
> that "top" displays. The normal "top" display on my machine shows
> "rc5des", "top", and "X" in the top three positions most of the time.
I will try to duplicate your problem here. I have 2.3.12-pre1 up for about
11 hours here on an SMP system running distributed-net and bonnie
filesystem benchmarks (usually run two at a time on the same filesystem to
try to break something). So far it is holding up pretty well. One thing I
have noticed is the distributed.net key blocks tend to vary in size and
therefore processing time:
[Jul 24 18:09:17 UTC] Completed one RC5 block B341F9EA:F0000000 (1*2^28 keys)
0.00:03:32.07 - [1,265,783.83 keys/sec]
[Jul 24 18:14:59 UTC] Completed one RC5 block B3350A61:50000000 (8*2^28 keys)
0.00:28:14.56 - [1,267,274.58 keys/sec]
So there I have one block that completed in three minutes and another in
28 but the blocksizes were very different (the number of keys/sec
processed were about the same). So the thing to look for is not time per
block but the number of keys per second.
You might want to try pre-2.3.12-1 since it is supposed to fix some known
problems with 2.3.11
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/