No problems, looks great (after 5 hours anyway). cache/swap balancing
looks fixed too! I can't force swap up to silly levels either, without
cache shrinking to something small.
Problem now seems to be I/O. Here's my numbers for some recent kernels:
-------Sequential Output-------- ---Sequential Input-- --Random--
-Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks---
MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU
2.2.14
500 5294 87.2 10799 9.3 4305 8.4 5991 90.0 10568 7.4 93.5 1.6
2.4.46
500 6013 98.4 9097 7.1 2686 3.4 5806 86.8 11155 8.0 83.6 1.4
2.3.99-pre3
500 5547 91.1 10134 7.7 2983 3.9 4993 73.7 11015 7.3 87.5 1.5
2.3.99-pre5
300 5936 97.6 9908 7.8 2922 3.8 4670 70.3 10624 7.4 98.4 1.5
2.3.99-pre6
300 3515 58.5 9059 11.8 1606 2.0 1643 24.8 2008 1.5 64.5 1.3
2.3.99-pre7-6
400 3716 61.0 7662 8.1 2476 3.4 3445 51.9 5817 4.2 80.0 1.4
2.3.99-pre7-6-classvm
400 5556 91.0 7945 6.5 2711 3.3 4623 69.4 9668 6.9 88.6 1.3
400 5667 92.8 7977 6.8 2726 3.6 4611 69.1 9583 7.3 87.7 1.6
Well, the numbers speak for themselves. I get a simular drop for dbench
too.
Lawrence
-
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/
This archive was generated by hypermail 2b29 : Sun May 07 2000 - 21:00:20 EST