Shell Programming and Scripting

Alright, I have this script that pulls files from a few locations, process those files, creates a zip file, rsync's it and then removes everything.

The problem that I'm having is that I do not know how large the rsync'ed zip file is going to be. Right now I'm using a sleep command before I remove all the files and directories created, however I'm uncomfortable in leaving it like this. All it would take would be a large file that it needs to rsync and the script would delete the file before it's done being transferred.

So, is there a way of having the script check the progress of rsync? If not, how would I go about doing something along these lines?

If this is a Bash shell script you could try the 'wait' command.

See more here - Job Control Commands

Hope this helps.

I like to use a lock file. This way the script will loop till lock file is gone than re execute:

Code:

#
# create name for lock file
#
lockDir="/root/lock_files"
lockFilePath="$lockDir/file.lock"
#
#
#
# Loop through servers until file is no longer exists
#  
while [ -e "$lockFilePath" ]
do

   exit
done
#
#
#
# create new lock file
#
touch $lockFilePath
#
#
#
# loop through servers and sync docroots and code
#
/usr/bin/rsync  --stats -e ssh -rlvtgoDz --delete  /dir/dir/dir/ root@server:/dir/dir/dir/ &
#
#
#
# Remove lock file........
#
rm -f $lockFilePath

Be sure to set up ssh keys. This is run on a Red Hat system.

Good Luck

For some reason the wait command doesn't work according to the documentation that you gave me. I've tried just doing this...

Code:

rsync -urza -e "ssh -i rsync-key -l username -p xxxx" /filepath/filename.zip webserv:/filepath/

wait

And it doesn't wait at all.

As for the lockfile, I'm not sure what that script does, exactly.

Code:

lockDir="/root/lock_files"
lockFilePath="$lockDir/file.lock"

while [ -e "$lockFilePath" ]
do

   exit
done

touch $lockFilePath

/usr/bin/rsync  --stats -e ssh -rlvtgoDz --delete  /dir/dir/dir/ root@server:/dir/dir/dir/ &

rm -f $lockFilePath

As far as I can tell, it starts out telling where the lock directory and file is. Then you say if the locked file exists, then if it does, then exit the script. Then if it isn't there, then you touch the file (thereby creating a blank file, if my "touch" knowledge is correct), and then starts the rsync in a background process. After it starts making the rsync, it then removes the lock file.

I could see this somewhat working if you made the script call itself before the while statement, but I don't know how this would improve my situation. Normally my script then removes the files that are being rsync'ed, so even if I called this script externally, the rm command would go through just as soon as this script finished, a la just as the rsync starts. And I really don't need to rsync multiple files, just one.

Any other ideas / suggestions?

(Thanks for what I have so far, by the way. )

EDIT: Also, I'm running all this on a Red Hat system. Forgot to include that earlier.

If I understand this correctly, you're looking for a way to make sure that the rsync completes successfully before moving on to the rest of your script. bash would allow you to do this with the '&&' operator (in this case rsync is command1 ):

Code:

command1 && command2

From the bash man page, "command2 is executed if, and only if, command1 returns an exit status of zero."

So, something you might be able to use would be :

Code:

rsync -options file.xx remotehost:/path/to/store && rm file.xx

The gotcha here is that rsync can fail for a couple different reasons and if it does, the second command would not execute.