I have a 4*Xeon 700A+2G memory server running reiserfs+2.4.18+openwall
patch.
Our service is running with large processes(> 4000) and a little heavy
disk I/O, and
the server's load is about 3.x to 4.x.
But last month after I increased 2G memory to 4G, the server is always
froze every two
or three hours. After 5-10 minutes, it can be alive again.
I have changed to RedHat kernel 2.4.18-17, and it has the same problem
but a little well--
it is froze only several seconds. And the average load is above
20,higher than 2.4.18.
Kernel build with 4G memory support + reiserfs + smp support + intel
e100 driver module.
Any advice? Thank for help.
2G + linux 2.4.18
# uptime
9:30pm up 4 days, 9:22, 7 users, load average: 3.48, 4.63, 4.63
# vmstat 2 2
procs memory swap io system
cpu
r b w swpd free buff cache si so bi bo in cs us
sy id
2 5 1 294264 20832 321096 584276 7 6 25 15 28 23 10
7 27
4 1 0 294196 13512 322424 585088 828 0 1476 540 7025 4090 18
31 51
4G + 2.4.18 freeze
# uptime
3:30pm up 3:09, 5 users, load average: 534.48, 154.42, 69.21
# vmstat 2 2
procs memory swap io system
cpu
r b w swpd free buff cache si so bi bo in cs us
sy id
276 14 8 0 43788 191968 2047620 0 0 82 157 822 404
12 31 57
6 30 2 0 14704 191156 2048776 0 0 760 3564 9212 3028 21
78 1# uptime
4G + redhat 2.4.18-17
#uptime
1:25pm up 1:04, 5 users, load average: 27.79, 23.25, 20.59
# vmstat 2 2
procs memory swap io system
cpu
r b w swpd free buff cache si so bi bo in cs us
sy id
1 1 1 0 664524 277408 1585192 0 0 113 140 740 398 11
22 68
0 1 2 0 664928 277852 1585428 0 0 240 1382 4470 2194 17
43 40
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Oct 31 2002 - 22:00:26 EST