[cobalt-users] Re: Raq4 - Error rotating logs with 1gb of ram!

["Yahoo" <jjma100@xxxxxxxxxxx>]

Hello

> that al looks good... try...
>  /usr/local/sbin/split_logs web < /var/log/httpd/access
> and see if you get an error...

Tried this manually and it crashed the server with problems in the kernel
logs. Rebooted the server, reran the split_logs script and it worked without
a problem.

Dec 23 00:32:32 l4i kernel: Unable to handle kernel paging request at
virtual address 2e2d9c50
Dec 23 00:32:32 l4i kernel: current->tss.cr3 = 1293f000, %%cr3 = 1293f000
Dec 23 00:32:32 l4i kernel: *pde = 00000000
Dec 23 00:32:32 l4i kernel: Oops: 0000
Dec 23 00:32:32 l4i kernel: CPU:    0
Dec 23 00:32:32 l4i kernel: EIP:    0010:[do_generic_file_read+64/1500]
Dec 23 00:32:32 l4i kernel: EFLAGS: 00010282
Dec 23 00:32:32 l4i kernel: eax: db815704   ebx: c0284040   ecx: db815740
edx: 080a0904
Dec 23 00:32:32 l4i kernel: esi: 0809cf40   edi: 00000000   ebp: 00008000
esp: fb245f3c
Dec 23 00:32:32 l4i kernel: ds: 0018   es: 0018   ss: 0018
Dec 23 00:32:32 l4i kernel: Process gzip (pid: 22362, process nr: 133,
stackpage=fb245000)
Dec 23 00:32:32 l4i kernel: Stack: 00000000 00008000 00000000 fa6c8000
fc8214d7 fa6ca4cc 00c71549 fa6c8000
Dec 23 00:32:32 l4i kernel:        fc821334 00000081 fb245f9c f9194220
c011d087 db815740 db815754 fb245f8c
Dec 23 00:32:32 l4i kernel:        c011cfd4 db815740 ffffffea 00000000
00000000 00008000 0809cf40 00000000
Dec 23 00:32:32 l4i kernel: Call Trace: [eepro100:eepro100_init+5139/13404]
[eepro100:eepro100_init+4720/13404] [generic_file_read+$
Dec 23 00:32:32 l4i kernel: Code: 8b 1c 83 31 ed 8b 74 24 40 8b 4c 24 3c 8b
36 89 74 24 30 89
Dec 23 00:32:32 l4i kernel: Unable to handle kernel paging request at
virtual address acb9b444
Dec 23 00:32:32 l4i kernel: current->tss.cr3 = 0c5cd000, %%cr3 = 0c5cd000
Dec 23 00:32:32 l4i kernel: *pde = 00000000

> we might want to purge your access file
> how to purge... :)
> cat /dev/null > /var/log/httpd/access
> /usr/bin/killall -HUP httpd

Ok carried this out and will monitor the logrotate tomorrow night to see if
it fails.

> also
> df -h and post it...

This is my output:

[root Scripts]# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/md1              2.8G  1.1G  1.7G  38% /
/dev/md3              1.9G   15M  1.8G   1% /var
/dev/md4               31G  2.4G   28G   8% /home

Hope this helps

Jonathan