Amanda-Users

Re: dump tape full?

2006-06-12 14:36:40
Subject: Re: dump tape full?
From: Kristian Rink <krink AT pm-planc DOT de>
To: amanda-users AT amanda DOT org
Date: Mon, 12 Jun 2006 20:28:26 +0200
Hello John, all:

and at first thanks for your comments...


Jon LaBadie schrieb:
>> to have completed without errors, but, running amverify ended up with
>> the following:
> 
> This is just guesswork, given that I'm bad and don't run amverify.

Hmmm, honestly I am also bad most of the time but a little more careful
right now - our backup system consists of an external SCSI-to-IDE RAID
system and an LTO-2 autoloader; hosts on the network dump their data to
the RAID system and amanda is about to hammer those data to tape at
night (incremental) or weekend (full). The RAID system hasn't been in
good condition lately after some of the drives went down, so I am
worried about the state of my tape backup, as well... That's why I
initially played around with amverify in order to see whether there
actually _is_ anything backed up at all... :)




>> amverify Full
>> Mo Jun 12 13:01:54 CEST 2006
...
>> Checked backer.planconnect.net.planc1.20060610.0
>> End-of-Information detected.
>> Loading next slot...
>> ...
>>> 
> Does the list of DLEs shown by amverify match the total of
> what were expected?

Yes. The way it looks until then, at least every DLEs have been
processed and found backed up on tape.




> Did the report that amdump mailed out report early on that
> "Some dumps may have been left on the holding disk"?

No, nothing like this. Looking at the report mails, everything is fine...





> The questions above would help confirm this guess, but I suspect that
> taper actually ran out of space on the tape.  Whether the DLE it was
> taping when it reached the end was backer.planconnect.net.planc1.20060610.0
> or the next DLE I don't know.  When taper hits the end of the tape the
> emailed report shows "successful tape usage", that is your 84%.  The
> remaining 16% might have been a partially taped DLE.


To be a little more verbose on that, here's an excerpt taken from
yesterdays Full dump log (sizes):


[...]
Output Size (meg)      172568.1   172568.1        0.0
Original Size (meg)    372550.6   372550.6        0.0
Avg Compressed Size (%)    46.3       46.3        --
Filesystems Dumped            7          7          0
Avg Dump Rate (k/s)      1198.0     1198.0        --
[...]
Tape Size (meg)        172568.1   172568.1        0.0
Tape Used (%)              86.4       86.4        0.0
Filesystems Taped             7          7          0
[..]
USAGE BY TAPE:
  Label        Time      Size      %    Nb    Nc
  Back00      40:59 176709760k   86.4     7     0


I'm not sure right now, but the way I understand this, we did back up
372.5 gig of data (which seems reasonable looking at the RAID disk
usage), compressed to 172.5gig. Assuming that LTO-2 at this level is
capable of storing 200gig data, the 86.4% usage of that very (single)
tape do make sense.

This way, the "End-Of-Information" message seems rather obscure... :o
Any more ideas, anyone? :)

Thanks for your patience and bye,
Kris





-- 
Kristian Rink   -- Programmierung/Systembetreuung
planConnect GmbH * Strehlener Str. 12 - 14 * 01069 Dresden
0176 24472771 * krink AT pm-planc DOT de

<Prev in Thread] Current Thread [Next in Thread>