Amanda-Users

Re: Using amrecover - strange happenings

2005-10-01 10:45:28
Subject: Re: Using amrecover - strange happenings
From: Jon LaBadie <jon AT jgcomp DOT com>
To: amanda-users AT amanda DOT org
Date: Sat, 1 Oct 2005 10:22:49 -0400
On Fri, Sep 30, 2005 at 03:43:41PM -0700, Steve H wrote:
> Ok, new error.  I was going to use dd to get the image off, and try to restore
> it in other ways.  Unfortunately, that didn't work, I received this error:
> 
> [root@hostname restore]# dd conv=noerror if=/dev/nst0 bs=32 skip=1
> of=./restoreimage
> dd: warning: working around lseek kernel bug for file (/dev/nst0)
>   of mt_type=0x72 -- see <sys/mtio.h> for the list of types
> dd: reading `/dev/nst0': Cannot allocate memory
> 0+0 records in
> 0+0 records out
> 

'cannot allocate memory' has (in my experience) been dd's cryptic
way of saying "the tape blocks are bigger than my input block size".

This is not surprising as you told dd to use a '32 byte' block size
and amanda uses a minimum of '32 kilobyte' blocks.

Try bs=32k.  But check the man page for the suffix.  I think k is 1024
and kb is 1000.

If that gives a similar message, it may be that your tape was written
with larger or variable block sizes.  So give another try with separate
input and output block sizes, say ibs=1m (or 16m) and obs=some smaller
size.  The latter is my attempt to avoid huge padding on a possible
last, partially-filled block.  Otherwise just use bs=1m (or 16m).

-- 
Jon H. LaBadie                  jon AT jgcomp DOT com
 JG Computing
 4455 Province Line Road        (609) 252-0159
 Princeton, NJ  08540-4322      (609) 683-7220 (fax)

<Prev in Thread] Current Thread [Next in Thread>
  • Re: Using amrecover - strange happenings, Jon LaBadie <=