I have a 486DX4/120 with 40MB of RAM running Red Hat 6.1. I've just installed
2.3.99-pre7, having followed most of the 2.3.99-pre series.
I know this has been discussed, but here's another data point; VM
performance is terrible.
At the moment, "free" gives:
total used free shared buffers cached
Mem: 38696 29628 9068 0 172 9636
-/+ buffers/cache: 19820 18876
Swap: 133016 1756 131260
...so I have plenty of spare RAM. This is running KDE but not much else. I just
started kmail, and it took much longer than previously to load, with much disk
activity. kswapd sat at the top of the "top" listing for a couple of refresh
cycles (about 8s each!) with CPU usage of 30-60%. The actual amount of swap
used changed by about 200KB after kmail had loaded.
I know I don't have a monster CPU, but the amount of disk activity seems
disproportionate to the task in hand.
I've just tried loading Netscape, and it's still tryng to load as I write! A
snapshot of "top" shows:
2 root 10 0 0 0 0 SW 0 42.7 0.0 1:09 kswapd
661 root 5 0 2044 4 2040 S 0 8.4 5.2 0:59 X
719 mark 12 0 348 4 344 R 0 3.7 0.8 0:31 top
763 mark 0 0 576 0 576 S 0 2.1 1.4 0:00 netscape-comm
737 mark 1 0 2024 92 1932 S 0 1.9 4.9 0:18 kmail
746 mark 0 0 716 0 716 D 0 1.8 1.8 0:01 ld-linux.so.2
Now Netscape has loaded, I still only have 5.5MB swapped and 16MB cache. kswapd
has taken over 1min 30s of my 25min uptime! I'm going back to 2.3.99-pre6.
-- Mark Tranchant mark@tranchant.freeserve.co.uk- 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 : Sat May 13 2000 - 04:45:55 EDT