Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
/ 2004-07-06 20:58:20 -0300
\ Marlon Cabrera Oliveira:
> >there should have been a complete oops after this?
> >pipe it through ksymoops, and post it.
>
>
> Unable to handle kernel paging request at virtual address 8b73af8c
> c01f99bd
> *pde = 00000000
> Oops: 0000
> CPU: 0
> EIP: 0010:[can_dfsa_file+61/384] Not tainted
^^^^^^^^^^^^^
> EFLAGS: 00210282
> eax: 00000000 ebx: 8b73af80 ecx: f2ed2000 edx: 0000000a
> esi: 00000001 edi: 0000000a ebp: c02f4480 esp: f2ed3f58
> ds: 0018 es: 0018 ss: 0018
> Process kdm_greet (pid: 1179, stackpage=f2ed3000)
^^^^^^^^^
> Stack: 00020572 00000000 001e7f87 00000000 0024ee46 f2ed2000 00000001 0000000a
> c02f4480 c01fa778 0000000a 0000000a 080c4ef8 fffffff7 00000001 c0148411
> 0000000a 00000001 bffff568 c01df215 f2ed2000 00000000 f2ed2000 080c4ef8
> Call Trace:
> [dfsa_syscall_on_file+376/400]
> [sys_write+49/352]
> [local_pre_usermode_actions+21/224]
> [local_syscall+7/19]
> Code: 8b 53 0c a8 02 74 2c 0f bf 4a 40 85 c9 75 0c 0f bf 4a 3e 85
> Using defaults from ksymoops -t elf32-i386 -a i386
since the call trace not even reached drbd,
it is very unlikely that this is a drbd issue.
> the clash occours to sync 20 GB reiserfs partition. the speed is too
> slow: 3,600 k/sec
>
> Sync rate was set with:
>
> drbdsetup /dev/nb0 syncer --min 599M --max 600M --nice -19
>
> I test drbd on a 1.9 G reiserfs partition, but the speed don' t increase.
>
> version: 0.6.12 (api:64/proto:62)
>
> 0: cs:SyncingAll st:Primary/Secondary ns:1895076 nr:0 dw:0 dr:1895104 pe:50
> ua:0
> [===================>] sync'ed: 97.2% (55/1906)M
> finish: 0:01:15h speed: 3,660 (3,318) K/sec
> 1: cs:Unconfigured st:Secondary/Unknown ns:0 nr:0 dw:0 dr:0 pe:0 ua:0
>
>
> Is the ATA disks 7200 rpm of 2 nodes the bootleneck??
>
> hdparm out:
>
> /dev/hda:
> multcount = 16 (on)
> IO_support = 1 (32-bit)
> unmaskirq = 0 (off)
> using_dma = 0 (off)
^^^^^^^^^^^^^^^^^^^^^^^
> keepsettings = 0 (off)
> readonly = 0 (off)
> readahead = 8 (on)
> geometry = 9733/255/63, sectors = 156368016, start = 0
>
> /dev/hda:
> Timing buffer-cache reads: 2048 MB in 2.00 seconds = 1024.00 MB/sec
> Timing buffered disk reads: 14 MB in 3.09 seconds = 4.53 MB/sec
^^^^^^^^^^^
> I setup dma only for disks on kernel.
and obviously it is not enabled nevertheless.
maybe your chipset is blacklisted because known
to corrupt data or hang or crash when dma is enabled.
or it had some disk problem, and did a dma reset meanwhile.
either way, there should be something hda related earlier in your syslog.
that oops is not a drbd problem. maybe it is more likely to occur
because drbd puts the system under stress...
check wether your chipset is properly supported.
run some hardware diagnostics.
maybe try a different debian or kernel.org kernel.
maybe try a different vendors kernel.
maybe try even different hardware.
Lars Ellenberg