Messages in this thread | | | Date | Sun, 29 Aug 1999 15:16:10 +0200 | From | Ward Vandewege <> | Subject | wait_on_bh trouble on SMP box with 2.2.11 |
| |
Hi,
I am having trouble with a dual PII-450Mhz box, with 4 SCSI disks (totalling 22 GB). It has 31.000 e-mail and web accounts (!), and currently runs vanilla 2.2.11. The machine is quite heavily loaded, with load averages from 2 to 4, depending on the time of the day. Every now and then, on average 2 to 3 times per day, it just locks up with the following on the console:
wait_on_bh; CPU 0: irq: 0 [0 0] bh: 1 [0 1] <[c010b24d]><[c014d9a6]><[c014d942]><[c012799f]><[c0128bbf]><[c0127a1a]> <[c0127a84]><[c010915c]>
The machine is locked solid. Doesn't respond to ping anymore. I've looked up these addresses in the System.map file:
<[c010b210]>: synchronize_bh <[c014d8f8]>: sock_poll <[c014d91c]>: sock_close <[c0127980]>: __fput <[c0128ba8]>: fput <[c01279c8]>: filp_close <[c0127a24]>: sys_close <[c0109128]>: sys_call
I have seen reports of this problem earlier this year on this list, with early 2.2.x kernels, but couldn't find a solution. Has anyone experienced this with 2.2.11?
The box used to run 2.0.36, and was stable - uptimes of months. But the load became too big, so we had to upgrade to a 2.2 kernel for better SMP support. The hardware has not been changed.
I would be most grateful for any comments,
Ward Vandewege.
- 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/
| |