Original message :
I don't have a DAT drive on the Personal Workstation 500. Is it
possible to do a vdump backup of the workstation over the network to
a tape device on a Alphaserver 2100A (DLT or DAT) ?
If so whats involved ? on the 2100A the DAT drive is /dev/rmt0h, can
I try to nfs mount it and then refer to the device as
/dev/rmt0h_at_2100_hostname ?? I don't think this would work ! >
Also, as an aside, can I do the same for a SUN Solaris workstation
?
----------------------------------------------------------------------
Thanks for all the prompt replys. There are several ways to skin a
cat, but the solution I have gone with is vdump/vrestore piped into
rsh.
First, I modified the .rhosts file on the server to allow root on the
workstation to connect without a password.
Then, from the workstation
vdump -0 -F 64 -b 60 -f - / | rsh juliet dd ibs=60k obs=60k
of=/dev/rmt0h
Simple really ! This worked fine,it was fairly slow. We are looking at
ways of speeding this up across the network by changing the IP address
of the workstation so that its on the same backbone as the
AlphaServer.
Now all I need to do is test that the restore works ! .
Other solutions/ideas mentioned include gtar and rdump/rrestore for
ufs,
Thanks to
T o m L e i t n e r
hl_at_tekla.fi
Jason.
Mark R Huehls
Alan
Lucia Gonzalez
Robert L. McMillin
Regards
Patrick
Below are the replies.
----------------------------------------------------------------------
Sure. The trick is to pipe the vdump output over the network to the
other workstation. I use something like this:
# vdump -v -0 -F 64 -b 60 -f - /home |
rsh remote_host dd ibs=60k obs=60k of=/dev/rmt0h
provided that you have "rsh" access to the other workstation as root.
Another way would be to use "gtar" with
gtar cvf remote_host:/dev/rmt0h /home
Hope this helps -- Tom
T o m L e i t n e r
----------------------------------------------------------------------
vdump 0f - /"dir" | rsh a2100 "dd of=/dev/rmt0h"
Of course you need to set up .rhosts first.
> Also, as an aside, can I do the same for a SUN Solaris
workstation ? Sure. But the tapes have to be read back the same way to
the same machineplatform.
rsh a2100 "dd if=/dev/rmt0h" | ufsrestore xf -
But if you want a goo backup solution, have a look at DECnsr (i.e.
Legato Networker
for solaris). Your 2100 should a server license ready, all you need is
a client license
for you networked clients.
--harald
hl_at_tekla.fi
----------------------------------------------------------------------
If you have ufs filesystems you can do use rdump/rrestore -- look in
the man pages. Solaris supports the rdump command so I would use that
for the sun machine to the alpha box -- you may need to use some
options to support the different plateform but it should work.
If you are using AdvFS then you will have to use vdump, as your mail
states. You could nfs mount your filesystems on the remote box and
use vdump -- as vdump is filesystem independant you wont have any
trouble backing up a nfs. If you could arrange setting up some
/.rhosts (making your root a trusted user on the remote system) files
you could use a rsh command like this to do your backup (and the other
way round for the restore). Test it to your satisfaction before you
commit to the backup stratergy
vdump 0uf - | rsh remote_machine dd of=tape_device_file bs=32k
I think 32k is the output of the dump command.
If you have ufs filesystems you can do use rdump/rrestore -- look in
the man pages. Solaris supports the rdump command so I would use that
for the sun machine to the alpha box -- you may need to use some
options to support the different plateform but it should work.
If you are using AdvFS then you will have to use vdump, as your mail
states. You could nfs mount your filesystems on the remote box and
use vdump -- as vdump is filesystem independant you wont have any
trouble backing up a nfs. If you could arrange setting up some
/.rhosts (making your root a trusted user on the remote system) files
you could use a rsh command like this to do your backup (and the other
way round for the restore). Test it to your satisfaction before you
commit to the backup stratergy
vdump 0uf - | rsh remote_machine dd of=tape_device_file bs=32k
I think 32k is the output of the dump command.
Hope this helps,
Jason.
Actually patrick -- its 64k you want to use as 32k will probably give
you an error message like so...
vrestore: unable to read from device '/dev/rmt0h'; incomplete read
32768 < 65536
but 64k works a treat.
----------------------------------------------------------------------
set up network file system and do a remote backup. the 2100 would be
the remote device. setting up nfs is the real trick.
Huehls Mark R
----------------------------------------------------------------------
The common method for doing vdump to a remote tape is something like:
vdump <options>f - filesystem | rsh remote dd of=tape-drive
bs=block-size
You'll need the block to get dd(1) to write a reasonable block size to
tape and reblock whatever comes off the pipe. Then you'll need to
experiment with what options vrestore will need to get data it likes
so you can restore the backup. I've read of other using something
like this and not checking to see that they could restore until it was
too late.
re: NFS served devices.
You're right, it won't work. NFS serves file systems. When a special
file is served via NFS, the local host interprets the device number.
Alan
----------------------------------------------------------------------
We had the same problem. We have 9 AlphaStations 255 without local
tape units, and an AlphaServer 8400 with a DLT and a DAT. We used to
dump everything on a large file on each workstation (with vdump) and
then just copied the 9 files to a NFS exported directory on the 8400
(/backup), from which we just did a
tar -cvf of /backup to the /dev/nrmt0h.
But this was disastrous, when someone wanted to recover a file, we had
to do a painful job of unpacking everything, blah, blah.
So we came up with a solution, a vdump|dd backup through rsh.
Steps:
1) on the AlphaStation console:
# /sbin/shutdown -h now (single user, we want the backup to be
perfect)
>> boot -fl s
# /sbin/bcheckrc (check and mount all filesystems)
# /sbin/init.d/inet start (so that the system knows its name) #
/usr/sbin/inetd /etc/inetd.cccfc.bck
where /etc/inetd.cccfc.bck is a file that just contains 1 network
service: rsh
----------------------------------------------------------------------
-
shell stream tcp nowait root /usr/sbin/rshd rshd
-l
----------------------------------------------------------------------
-
Thus we leave the machine accesible with rsh by root from the server,
but no other network services open, and no rsh for the rest of users
(flag -l)
2) On the AlphaServer console:
# mt -f /dev/nrmt0h rewind (or fsf #n if it's not the beginning of the
tape) # /backup/script.ksh
where /backup/script.ksh does the following (for level 0 backups)
----------------------------------------------------------------------
---------- ------------
#!/bin/ksh
echo "Which machine are you backing up?" read workstation
today=`date +%d/%m/%y`
logfile=/backup/$workstation.$today.log
time rsh $workstation "/sbin/vdump -0uf - -b 1 /" | dd of=/dev/nrmt1h
bs=1024 >> $logfile 2>&1
time rsh $workstation "/sbin/vdump -0uf - -b 1 /usr" | dd
of=/dev/nrmt1h bs=1024 >> $logfile 2>&1
mailx -s "Level 0 backup ($workstation/$today) to DLT tape"
root_at_mailaccount < $logfile
rm $logfile
----------------------------------------------------------------------
---------- -------------
it works! if you want to recover anything from the backup you just
have to sit on console on the server and:
# mt -f /dev/nrmt0h fsf #N
# cd /backup/recover
# "dd if=/dev/nrmt1h bs=1024"|vrestore -if -
you get the interactive prompt from vrestore and that's fairly easy...
But this strategy is incomplete, since we would like to have the
backups done during the night, causing the least disruption to the
users.
Ideally, there should be a rc.4 level on the workstation. This level
would start bcheckrc, inet, inetd, and kill everything else.
Then I would write a crontab on the AlphaServer, which sent at 2:00 AM
by rsh an
init 4 to the workstation, waited for a while, and then executed
/backup/script.ksh.
But I have no idea of how i can develop a level without doing anything
wrong! Any suggestions?
Thanks in advance,
Lucia Gonzalez
System Administration.
----------------------------------------------------------------------
> /dev/rmt0h_at_2100_hostname ?? I don't think this would work !
Nope, you'd say "vdump -f hostname:/dev/rmt0h /". However, that's
painfully slow unless you have 100BaseTX on all your machines. I
wouldn't recommend this for the folks at home anyway, since the rmt
protocol is slow to begin with once you consider overhead.
> Also, as an aside, can I do the same for a SUN Solaris
workstation ?
Yup, provided you have ~root/.rhosts correctly set up.
--
Robert L. McMillin
Received on Mon Oct 20 1997 - 17:55:07 NZDT