[comp.unix.ultrix] Problem with dump

bbraden@mesrx.UUCP (Bill Braden) (01/31/91)

Using dump under Ultrix V3.1 and trying to do a level 0 dump of /usr
I am getting the following result:
-------------------------------------------------------------------------
# dump 0sf 4800 /dev/nrmt0h /usr
  DUMP: Date of this level 0 dump: Wed Jan 30 08:47:29 1991
  DUMP: Date of last level 0 dump: the epoch
  DUMP: Dumping /dev/rra0c (/usr) to /dev/nrmt0h
  DUMP: Mapping (Pass I) [regular files]
  DUMP: Mapping (Pass II) [directories]

  DUMP: Estimates based on 4800 feet of tape at a density of 10240
BPI...
  DUMP: This dump will occupy 13296 (10240 byte) blocks on 0.39 tape(s).

  DUMP: Dumping (Pass III) [directories]
  DUMP: Dumping (Pass IV) [regular files]

		********************
  DUMP: (This should not happen)bread from /dev/rra0c [block 69872]:
count=8192, got=-1
		********************

  DUMP: 13295 tape blocks were dumped on 1 tape(s)

  DUMP: Dump is done
---------------------------------------------------------------------------
Apparently this should not happen, but it does.  Problem only occures on
level 0 dumps.  fsck finds no problems with the file system.  I am doing
this with the file system mounted (and probably active).  System is a
microvax II and the disk is an RD54, /usr is the only file system on the
disk and covers the entire disk.  Are these dumps worth the tape they
are written on? Is there a way to clear this?  

Please no lectures about dumping active file systems.  That subject has
already been covered.  Besides life is not without risk.

Thanks in advance for any help.

Bill Braden
!uunet!mesrx!bbraden		"Onward Through The Fog"
-------------------------------------------------------------------------
Any opinions expressed belong to me not Measurex.
-------------------------------------------------------------------------

vixie@decwrl.dec.com (Paul A Vixie) (02/01/91)

Sounds like a hard error.  What does "uerf" say?  Try bringing
the system down to single-user and using "radisk -s 0 -1 /dev/rra0c".

Fsck doesn't read data blocks, so it wouldn't find this.

Cheers,
--
Paul Vixie
DEC Western Research Lab	<vixie@wrl.dec.com>
Palo Alto, California		...!decwrl!vixie

rbj@uunet.UU.NET (Root Boy Jim) (02/01/91)

In article <529@mesrx.UUCP> bbraden@mesrx.UUCP (Bill Braden) writes:
>Using dump under Ultrix V3.1 and trying to do a level 0 dump of /usr
>I am getting the following result:
>  DUMP: This dump will occupy 13296 (10240 byte) blocks on 0.39 tape(s).
>  DUMP: (This should not happen)bread from /dev/rra0c [block 69872]:
>count=8192, got=-1
>  DUMP: 13295 tape blocks were dumped on 1 tape(s)
>  DUMP: Dump is done

OK, I won't flame you, but if you really want to know, you may have
to dump with the FS unmounted, or read only. What happens when you
dump the block device? Perhaps your FS is screwed up. Can you write
a C program (or use dd) to read block 69872? Can you take the tape
and restore it somewhere else (different machine or partition)?
You will note that only one block was missed. Use the duck test.
If your dump tape walks, talks, and quacks like a dump tape, then it's
problably a dump tape. Probably. As you say, life is not without risk.
-- 

	Root Boy Jim Cottrell <rbj@uunet.uu.net>
	Close the gap of the dark year in between

stevel@Autodesk.COM (Steve Litras) (02/06/91)

In article <529@mesrx.UUCP> bbraden@mesrx.UUCP (Bill Braden) writes:
>  DUMP: (This should not happen)bread from /dev/rra0c [block 69872]:
>count=8192, got=-1

We've had this problem too. In fact we just replaced the disk, so I don't know
how it will help (I'm assuming it will fix it, but you never know). According
to our Sun engineer, it's is a fairly harmless problem (soft error), but I
have had it bogus backups.