Re: Problems with amrecover/index/amcheck after tape problems

[Chr. Rossmanith]

I've pressed the "send" button too quickly and sent the answer to John only - here
is a copy for the mailing list...

Christina


> Hi John,
>
> "John R. Jackson" wrote:
>
> > >This is the content of our tapelist file. I think it should have more than
> > >one line?!?
> > >
> > >20010725 NEURO007 reuse
> >
> > Ooops.  You've lost your tapelist file, which is a very bad thing.
> > That's why Amanda is asking about new tapes and complaining about backups
> > being overwritten.
> >
> > It's also why I make a backup of it (and a lot of other critical Amanda
> > files) before each run and save several copies, just in case.
>
> Is there a list of those critical files? And maybe a description which files are
> used for which purpose? Otherwise I just don't know what to backup manually
> before running amanda...
>
> > You may be able to put it back together, though.  I just tried the
> > following ksh code.  It finds the taper START line in each log.MMDD.NN
> > file and rebuilds the tapelist file from that.
> >
> >   rm tapelist.log
> >   cat log.* | grep '^START taper' | while read x x x datestamp x label x
> >   do
> > echo $datestamp $label reuse >> tapelist.log
> >   done
> >   sort -rn < tapelist.log > tapelist.new
> >
> > At this point, you will hopefully have 25 tapes listed in tapelist.new.
> > Look through it and make sure thing seem right.  In particular, make
> > sure a tape is not listed twice.
>
> With this script I've gathered the information from /var/log/amanda/NEURO and
> .../oldlog together and got a tapelist.new which was too long (104 lines) and
> removed the oldest entries and now I have a nice tapelist again  :-)
> Your great...
>
> > Check your current tapelist file.  It should be owned by your Amanda user
> > and mode 0600.  Move it out of the way and copy tapelist.new to tapelist,
> > setting the ownership and mode.
> >
> > Finally, try "amadmin  tape" again and see if it's happier.
>
> ...yes, it's singing and smiling all the time and requests tape NEURO009 for
> tonight (as expected  :-)   )
>
> Just amrecover is complaining about missing logfiles. The date 2001-07-19, which
> I'm interested in, isn't affected because this logfile exists. I think I could
> move the missing logfiles, which have obviously been moved too early, from oldlog
> to the log directory...
>
> Now amrecover finished successfully  :-)
>
> Thanks a lot for your help. Using amrecover is much more fun than using
> amrestore...
>
> Christina Rossmanith

Re: Problems with amrecover/index/amcheck after tape problems

[Chr. Rossmanith]

Hi John,

thank you for your quick response. This night the backup was finished
successfully but it tells a lot of overwriting backups on previous tapes.
I've attached just the NOTES section from tonights mail:

NOTES:
  planner: Last full dump of neuroserver:/home2 on tape
NEURO001 overwritten in 1 run.
  planner: Last full dump of neuroserver:/home3 on tape
NEURO003 overwritten in 1 run.
  planner: Last full dump of neuroserver:/var/spool/mail on
tape NEURO002 overwritten in 1 run.
  planner: Last full dump of neuro021:/var/lib/pgsql on tape
NEURO002 overwritten in 1 run.
  planner: Last full dump of neuro034:/home on tape NEURO002
overwritten in 1 run.
  planner: Last full dump of neuro057:/home on tape NEURO002
overwritten in 1 run.
  planner: Last full dump of neuro057:/usr on tape NEURO001
overwritten in 1 run.
  planner: Last full dump of neuroserver://neuro068/HeineT on
tape NEURO002 overwritten in 1 run.
  planner: Last full dump of neuroserver://neuro031/docu on
tape NEURO002 overwritten in 1 run.
  planner: Last full dump of neuroserver://neuro040/dbase on
tape NEURO002 overwritten in 1 run.
  planner: Last full dump of neuroserver://neuro040/liquid on
tape NEURO002 overwritten in 1 run.
  planner: Last full dump of neuroserver://neuro040/texte on
tape NEURO002 overwritten in 1 run.
  planner: Last full dump of neuroserver://neuro040/priv2000 on
tape NEURO002 overwritten in 1 run.
  planner: Last full dump of neuroserver://neuro065/epilepsie
on tape NEURO002 overwritten in 1 run.
  planner: Last full dump of neuroserver://neuro035/user$ on
tape NEURO001 overwritten in 1 run.
  planner: Last full dump of neuroserver://neuro027/schneider$
on tape NEURO002 overwritten in 1 run.
  planner: Last full dump of neuroserver://neuro055/genetic on
tape NEURO002 overwritten in 1 run.
  taper: tape NEURO007 kb 11427200 fm 23 [OK]


And for recovering a file from tonights backup I can use amrecover:

This is what amrecover tells me after start up and changing to the disk in
question:

neuroserver# ~backup/sbin/amrecover
AMRECOVER Version 2.4.2p1. Contacting server on neuroserver ...
220 neuroserver AMANDA index server (2.4.2p1) ready.
200 Access OK
Setting restore date to today (2001-07-25)
200 Working date set to 2001-07-25.
200 Config set to NEURO.
200 Dump host set to neuroserver.
Can't determine disk and mount point from $CWD
amrecover> setdisk /opt/samba2.1
Scanning /opt/dumps...
200 Disk set to /opt/samba2.1.

BUT: setting the date to something earlier gives:

amrecover> setdate 2001-07-12
200 Working date set to 2001-07-12.
No index records for cwd on new date
Setting cwd to mount point


"John R. Jackson" wrote:

> >I'd like to recover a file from the backup using amrecover.
> >Unfortunately amrecover tells me, that there is no index available for
> >the given disk/date/host. BUT: As fas as I can see, there are all index
> >files in the index directory /usr/adm/amanda...
>
> OK, let's tackle this problem first.
>
> It would help to see the actual amrecover messages (all of them), as
> well as know what version of Amanda you're using.

This is included in the amrecover message above...

> First, let's make sure the index files are where they should be.  Run:
>
>   amgetconf  indexdir

$ amgetconf NEURO indexdir
/usr/adm/amanda/NEURO/index

> That points to the top level of the index directory.  Within there should
> be a directory for each of your clients.  The name will be the same as
> (or very close to -- '/' and whitespace are converted to '_' and '_'
> is converted to '__', but that's rare in a host name) the disklist entry.
>
> Within each client directory will be a directory for each disk.  Those
> names will be similar (see above) to the disk name, e.g. "/dev/hda5"
> would be "_dev_hda5".
>
> Within a disk directory will be the gzip'ed index files.  They have
> a datestamp and level in the name, and you must have all of them back
> through the full dump preceeding the date you're requesting.

The directory content of indexdir/neuroserver/_opt_samba2.1 is

20010614_0.gz  20010622_1.gz  20010630_2.gz  20010710_0.gz  20010718_0.gz
20010615_1.gz  20010623_1.gz  20010703_0.gz  20010711_1.gz  20010719_1.gz
20010616_1.gz  20010626_0.gz  20010704_1.gz  20010712_0.gz  20010720_2.gz
20010619_2.gz  20010627_1.gz  20010705_1.gz  20010713_1.gz  20010721_2.gz
20010620_0.gz  20010628_1.gz  20010706_2.gz  20010714_1.gz  20010725_0.gz
20010621_1.gz  20010629_2.gz  20010707_2.gz  20010717_1.gz

Backups on Monday and Tuesday failed -> index files from 2001-07-23 and
2001-07-24 are missing. If I understand you right, setting the date to
2001-07-19 should be successful, because the last full dump is from 07-18.
But this fails with the same error message as setting the date to 07-12
(amrecover message above).

> Also, if you're using GNU tar, make sure the files are formatted properly.
> If you look at the first few lines and they start with a large number, it
> means you're using a broken version of GNU tar.  You'

Problems with amrecover/index/amcheck after tape problems

[Chr. Rossmanith]

Hi,

I'd like to recover a file from the backup using amrecover.
Unfortunately amrecover tells me, that there is no index available for
the given disk/date/host. BUT: As fas as I can see, there are all index
files in the index directory /usr/adm/amanda...

We had some problems this week because the backup server could not talk
to the tape drive (mt status failed as well) and all backups stayed on
the holding disk. The problem is fixed right now (just by waiting), I
performed an amflush and thought that now everything should go one like
before...

The backups were flushed to the tape NEURO006 and I would have expected
that amanda requests tape NEURO007 next, but it tells me that it would
like a new tape. Am I right that I would have been informed if one tape
were not enough for amflush?!?

Something else is strange: after amflush I tried amcheck still having
tape NEURO006 in the tape drive. And amcheck was happy. It was happy
with NEURO007 as well.

Could the problems accessing the tape drive during the last days have
left some corrupt info files?!?

I would really be glad to get some advice where to look and what to do!


Thank you,

Christina Rossmanith

amrecover fails on client

[Chr. Rossmanith]

Hi,

when extracting files from a backup on a client using amrecover I get
the following error message:

amrecover: error reading tape: Connection reset by peer
extract_list - child returned non-zero status: 1

Using amrecover on the server works fine. Any hints what might go
wrong?!?

Configuration:

Server: Linux, amanda 2.4.2p1
Client: Linux, amanda 2.4.2p1

Any information missing? Just ask, I'll provide them...


Thank you,

Christina Rossmanith