装了2台服务器，都有点问题

[刘世伟 <liushiwei@anheng.com.cn>]

装了2台服务器，都有点问题
2台服务器都是amd64 X2 4200+的cpu 但主板不同。
系统偶尔会崩溃。 崩溃是交换内存被占满， 机器能回应ping,但是所有的服务都停止响应。
崩溃时kern.log的提示完全一样,估计是交换内存占满造成的错误提示,
会不会是2.6.18-3-amd64的问题？
现在换了kernel.org下载的2.6.20-1内核， 正在测试中。

其他人有遇到这个问题的吗？ 


出错是的kern.log提示:
Feb 24 03:15:49 YZ kernel: Call Trace:
Feb 24 03:15:49 YZ kernel:  [<ffffffff802a7134>] out_of_memory+0x33/0x218
Feb 24 03:15:49 YZ kernel:  [<ffffffff8020e094>] __alloc_pages+0x220/0x2a9
Feb 24 03:15:49 YZ kernel:  [<ffffffff80210e78>] 
__do_page_cache_readahead+0x95/0x206
Feb 24 03:15:49 YZ kernel:  [<ffffffff80226998>] sync_page+0x0/0x42
Feb 24 03:16:37 YZ kernel:  [<ffffffff8025d501>] io_schedule+0x28/0x33
Feb 24 03:16:53 YZ kernel:  [<ffffffff80226998>] sync_page+0x0/0x42
Feb 24 03:17:19 YZ kernel:  [<ffffffff8025d798>] __wait_on_bit_lock+0x5b/0x66
Feb 24 03:17:30 YZ kernel:  [<ffffffff8023dc39>] __lock_page+0x5e/0x64
Feb 24 03:17:30 YZ kernel:  [<ffffffff802117b4>] filemap_nopage+0x148/0x314
Feb 24 03:17:30 YZ kernel:  [<ffffffff80208718>] __handle_mm_fault+0x375/0x94f
Feb 24 03:17:30 YZ kernel:  [<ffffffff8020a6d1>] do_page_fault+0x39d/0x706
Feb 24 03:17:30 YZ kernel:  [<ffffffff8020b03e>] vfs_read+0x13c/0x171
Feb 24 03:17:30 YZ kernel:  [<ffffffff802592dd>] error_exit+0x0/0x84
Feb 24 03:17:30 YZ kernel:
Feb 24 03:17:30 YZ kernel: Mem-info:
当时cpu温度:

03:08:00
CPU Temp:  +45.0掳C  (high = +80.0C, hyst = +75.0C)
SyS Temp:  +42.5C  (high = +80.0C, hyst = +75.0C)
03:14:00
CPU Temp:  +53.5C  (high = +80.0C, hyst = +75.0C)
SyS Temp:  +51.0C  (high = +80.0C, hyst = +75.0C)
03:14:00
CPU Temp:  +52.5C  (high = +80.0C, hyst = +75.0C)
SyS Temp:  +51.0C  (high = +80.0C, hyst = +75.0C)
03:15:00
CPU Temp:  +47.0C  (high = +80.0C, hyst = +75.0C)
SyS Temp:  +45.5C  (high = +80.0C, hyst = +75.0C)