Re: [9fans] worm problems

Re: [9fans] worm problems

[Russ Cox]

Can you boot a Plan 9 terminal on the disk
(or boot the install floppy) and then look
at the second block on your pseudo-worm?

Also, backups are good.

I had a problem like this once right after I'd
set up a pseudo-worm and other partition on the
same disk using the p device.  I never tracked
it down, but my superblock was the beginning of
a PostScript file.  The case fan had broken, so
I think it might have just been heat problems.

Russ

Re: [9fans] worm problems

[Sam Ducksworth]

russ,

i was able to boot from the install disk and do the following:

dd -if /dev/sdC0 -of /tmp/dump1.dd -skip 1 -count 1

and it wrote the result out to the file without any problems.

i did notice something strange. i mounted the file server on
my cpu server on /n/fs and then cd'ed into that directory and did
and 'lc' it had double entries for every directory. if i went into
one of the directories and did and 'lc' i did not see double entries.
it was only at the root of the mount point.

if it makes it any easier, i don't care about any of the previous dumps.

--sam

On Sun, 9 Dec 2001, Russ Cox wrote:

> Can you boot a Plan 9 terminal on the disk
> (or boot the install floppy) and then look
> at the second block on your pseudo-worm?
>
> Also, backups are good.
>
> I had a problem like this once right after I'd
> set up a pseudo-worm and other partition on the
> same disk using the p device.  I never tracked
> it down, but my superblock was the beginning of
> a PostScript file.  The case fan had broken, so
> I think it might have just been heat problems.
>
> Russ
>

--sam

Re: [9fans] worm problems

[Sam Ducksworth]

russ,

ok here is what i get  with that command:
....
0000ff0 02 00 00 00 05 00 00 00 00 00 01 00 02 00 00 00
0001000
0001000

the last six bytes are just as you said.

--sam

On Sun, 9 Dec 2001, Russ Cox wrote:

> try
> 	dd -bs 4096 -count 1 -skip 2 < /dev/sdC0/data | xd -b | tail -10
>
> to get the end of the superblock, where the "header" is.
> if i'm reading the code right, the last six bytes should be
> 	01 00 02 00 00 00
>
> the 01 00 is 0x0001 - the Tsuper tag; 02 00 00 00 is the
> block address: 2.
>
>

--sam

Re: [9fans] worm problems

[Russ Cox]

try
	dd -bs 4096 -count 1 -skip 2 < /dev/sdC0/data | xd -b | tail -10

to get the end of the superblock, where the "header" is.
if i'm reading the code right, the last six bytes should be
	01 00 02 00 00 00

the 01 00 is 0x0001 - the Tsuper tag; 02 00 00 00 is the
block address: 2.

[9fans] worm problems

[Sam Ducksworth]

hello all,

i noticed this morning that my fileserver was caught in a reboot cycle.
each time it comes up everything looks normal until just before it
displays the fs console prompt.

here is what it prints:

fworm: read 5
WORM SUPER BLOCK READ FAILED
0 0 fs:

then lots of disk activity followed by the fileserver panicing and
vomiting on the screen and then rebooting. i was able to stop the reboot
cycle by issuing the following command:

0 0 fs: cwcmd startdump 0

that got me back so i can at least use the file server. i also verified
the problem by issuing the following command:

0 0 fs: cwcmd touchsb
fworm: read 5
WORM SUPER BLOCK READ FAILED

any ideas on how to fix this one? BTW this is a pseudo worm setup.

--sam