Looks its going to be work.
I will update u shortly

Please note:
I do not want to create backup on /root on the failuer.


Thanks

You still don't need that last line. You've already performed the cd inside the IF section, you don't need to do it again. (This is not going to stop your script from working, I'm just removing unnecessary statements.)

Is it working now?

Yes, it is working.

But right now we do not encounter any mounting issue.
Script has worked.

I'd like to test the script.
How should I test?

Please note:
In the past on failure TAR backup has created on /root FS.
We do not want to create backup on /root.
I think this script will prevent to do TAR backup on /root FS.
It will create TAR backup only on defined path in the script.

Is it correct?

Yes, I'm confident that the script will terminate if it can't find the directory it wants to put the archive in.

If you want to test it, try these things:-

1. Comment the "tar" command
   Firstly, put a "#" at the start of the tar command, as well as any commands that you DON'T want executed (perhaps the mail commands as well)
2. Put the script in "debug" mode
   At the top of the script, put the command "set -x". This will cause the script to show you each command as it runs.

My Testing Results:

Script:

set -x
umount /mnt/upgdp
mount 00.00.00.0:/bkup /mnt/upgdp
if [ -d /mnt/upgdp/dev2 ]; then
# The preferred directory exists, so use it
cd /mnt/upgdp/dev2
else
echo Mount is not available, try agin later | mutt -x -s "Backup Failed" Mail@mail.com
exit 1
fi
cd /mnt/upgdp/dev2
rm dev2.bk*
today=$(date '+%m%d%y')
#tar -czvf dev2.bk_"$today".tar.tgz /opt/IBM/WebSphere
/home/dn/TARStatus > /home/dn/TARStatus.log
#mail -s "TAR Backup for dev2 has been done" Mail@mail.com < /home/dn/TARStatus.log


Results:

++ umount /mnt/upgdp
++ mount 00.00.00.0:/bkup /mnt/upgdp
++ '[' -d /mnt/upgdp/dev2 ']'
++ cd /mnt/upgdp/dev2
++ cd /mnt/upgdp/dev2
++ rm 'dev2.bk*'
rm: cannot remove `dev2.bk*': No such file or directory
+++ date +%m%d%y
++ today=032510
++ /home/dn/TARStatus

is it ok?

Many thanks for all of your help.

D

Can you help me on following too?

http://www.linuxquestions.org/questi...ep-w-e-797568/

Thanks

Yes, it all looks correct. Remove the test changes you made (i.e. set -x and the comments) and you can then use the script.

We have two backup servers.

upgdp (in script above) and bigbkp

Now users wants instead of backup failed message, if mount is not available on backup server1 (upgdp), it should place to backup server2 (bibbkp).

If both are not available then we get the backup failed message.


Could you correct below? thanks

umount /mnt/upgdp
mount 00.00.00.0:/bkup /mnt/upgdp
if [ -d /mnt/upgdp/dev2 ]; then
# The preferred directory exists, so use it
cd /mnt/upgdp/dev2
else
mount 00.00.00.0:/bigbkup/data /mnt/bigbkup
if [ -d /bigbkup/dev2 ]; then
cd /mnt/bigbkup/dev2
rm dev2.bk*
today=$(date '+%m%d%y')
tar -czvf dev2.bk_"$today".tar.tgz /opt/IBM/WebSphere
/home/dn/TARStatus > /home/dn/TARStatus.log
mail -s "TAR Backup for dev2 has been done on Server2" Mail@mail.com < /home/dn/TARStatus.log
exit 1
fi
else
echo Mount is not available, try agin later | mutt -x -s "Backup Failed" Mail@mail.com
exit 1
fi
cd /mnt/upgdp/dev2
rm dev2.bk*
today=$(date '+%m%d%y')
tar -czvf dev2.bk_"$today".tar.tgz /opt/IBM/WebSphere
/home/dn/TARStatus > /home/dn/TARStatus.log
mail -s "TAR Backup for dev2 has been done" Mail@mail.com < /home/dn/TARStatus.log

Try this code:-

Do you think its normal behaviour?

I umount upgdp as

umount /mnt/upgdp

but still I can go to the following directory.

cd /mnt/upgdp/wps2
> /mnt/upgdp/wps2

Actually I am testing new script , I am thinking to umount upgdp and put # b4 mount 00.00.00.0:/bkup /mnt/upgdp and run the script.

But in this case script will pick cd /mnt/upgdp/wps2 and do the backup on /mnt/upgdp/wps2.

Is it really umount?

thanks

OK, a quick overview of mount points.

A mount point - in this case /mnt/upgdp - is just a directory. Just like any other directory. If you create /mnt/upgdp as a directory, you can then put other data and subdirectories under it, just like a normal directory.

When you mount a file system on that mount point, the contents of the mounted file system are visible instead of the data and subdirectories you had there before. When you umount the file system, your existing data and subdirectories will "return". The fact is they didn't get deleted or lost, the mounted file system just takes precedence while it is mounted.

Does that explain what you're seeing?

When you have the file system mounted, you will see if in the output of the commands "df -k" or "df -h". When the file system is un-mounted, you won't see it using df because when the file system is un-mounted, the mount point is behaving like a normal directory entry and doesn't need to be in the "df" output.

This problem has been solved.

One of our team created the directory in the dev2 byitself.
I will test the script and update to you

from above after deleting dev2 directory from the server itself.
I have umount upgdp

Now I am getting this message.
I think this is expected.


cd /mnt/upgdp/dev2
-bash: cd: /mnt/upgdp/dev2: No such file or directory

No I will test script.

If dev2 exists in the remote file system, as I think it does, and you have unmounted the remote file system, then this is expected behaviour.

Many Thanks


All options are working.
Is it possible for you to look at my other request too?

http://www.linuxquestions.org/questi...ep-w-e-797568/



Thanks