Appache server hung after run out of memory

From: Hua Qin (qinhua@poisson.ecse.rpi.edu)
Date: Sat Sep 21 2002 - 10:12:51 EST


I got a Appache server hung, and the error_log showed on Appache server:
(12)Cannot allocate memory
(105)No buffer space available: write to cgi daemon process

Then I dropped this server to kdb and did backtrace:

PID: 23406 TASK: e0e70000 CPU: 1 COMMAND: "httpd"
   START: swap_out_pmd+234
e0e71d40: swap_out_vma+181
e0e71d78: smp_call_function_interrupt+50
e0e71d98: swap_out_mm+83
e0e71dc0: swap_out+184
e0e71de0: refill_inactive_zone+44
e0e71e04: refill_inactive+73
e0e71e24: do_try_to_free_pages+66
e0e71e40: try_to_free_pages+42
e0e71e54: _wrapped_alloc_pages+450
e0e71e74: __alloc_pages+18
e0e71e94: __get_free_pages+19
e0e71e9c: pte_alloc+151
e0e71ed8: copy_page_range+338
e0e71f04: __alloc_pages_limit+134
e0e71f1c: copy_mm+498
e0e71f64: do_fork+1269
e0e71f84: filp_close+158
e0e71fa8: sys_fork+22
e0e71fc0: system_call+51

I am wondering what will cause this swap_out_pmd hung? I did turn on the
swapon, and the swap patition size is 514072. I did not check how many
swap had been used when the hung happened.

The Appache version is 2.0 and Linux kernelis 2.4.7 (I know it's a little
bit old).

Thanks!
Hua

-
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 : Mon Sep 23 2002 - 22:00:33 EST