unusually high load averages
Jake Khuon
khuon at NEEBU.Net
Thu Jan 29 15:32:45 PST 2004
### On Thu, 29 Jan 2004 15:43:22 -0600, Jonathan Fosburgh
### <syjef at mdanderson.org> casually decided to expound upon
### freebsd-current at freebsd.org the following thoughts about "Re: unusually
### high load averages":
JF> Run `top` and see what your interrupts are doing. I'd guess you're suffering
last pid: 1011; load averages: 1.14, 0.93, 0.98 up 0+00:26:38
15:30:09
85 processes: 2 running, 73 sleeping, 10 waiting
CPU states: 0.8% user, 0.0% nice, 0.8% system, 0.4% interrupt, 98.1%
idle
Mem: 35M Active, 94M Inact, 38M Wired, 2480K Cache, 34M Buf, 74M Free
Swap: 1024M Total, 1024M Free
PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND
11 root -16 0 0K 12K RUN 22:39 98.83% 98.83% idle
661 root 96 0 42416K 33140K select 0:04 0.05% 0.05% XFree86
...
JF> from the IRQ20 storm. vmstat -i should also tell you that. A patch was
# vmstat -i
interrupt total rate
irq0: clk 158958 99
irq1: atkbd0 5423 3
irq6: fdc0 2 0
irq8: rtc 203476 127
irq11: cbb0 cbb1++* 91496 57
irq12: psm0 11877 7
irq13: npx0 1 0
irq14: ata0 49687 31
irq15: ata1 57 0
Total 520977 327
JF> posted to this list in the last couple of weeks that deals with this. It is
JF> working for me (I first ran into this a couple of days ago).
I tried to apply the patch but it didn't go in cleanly so I'll have to
hand-massage it. I did start running into this when I enabled ACPI because
the apm module is causing my kernel to panic for some reason. However I
also notice this when I have ACPI disabled too.
--
/*===================[ Jake Khuon <khuon at NEEBU.Net> ]======================+
| Packet Plumber, Network Engineers /| / [~ [~ |) | | --------------- |
| for Effective Bandwidth Utilisation / |/ [_ [_ |) |_| N E T W O R K S |
+=========================================================================*/
More information about the freebsd-current
mailing list