UNIX for Dummies Questions & Answers

Hey guys I'm trying to put a script together that will find files older than 30 days and move them to a backup location. Once moved I would like the files deleted from the "prior" directory. I also have it set to email me the status of if the files moved or no. I'm definately not a programmer of any sorts but below was the command I'm using. I put this together from reading some forums on here.

cp -r -p `find /edi/aiprod/report -mtime +30` /edilogs/aiprodreport
ERRCODE=`echo $?`
if [ $ERRCODE != 0 ] ; then
echo "Failed" | mailx -s "EDI Report Maintenance" xxxxxx@xxxxx.com
exit
else
echo "Successful" | mailx -s "EDI Report Maintenance" xxxxxx@xxxxx.com

rm -r `find /edi/aiprod/report -mtime +30`
fi


The problem I'm having with this is that when the script runs it doesnt keep the same directory structure. It seems like any file in a sub directory does not get copied over. Also the files that get copied over are put into the main directory it won't creat the sub directory the file was copied from. Does anyone have any suggestions on a better approach?

For ex.
/edi/aiprod/CATERPILLAR/20081216.850.doc

would be copied to

/edilogs/aiprodreport/20081216.850.doc

Hopefully I made some sense of this. Any help would be much appreciated thanks a ton guys!

I think a better tool for the job is rsync.

rsync --files-from=<(find /../src -mtime +30 -print) src/ dest/

there are also rsync files to delete afterwards.

You can also use tar -c | tar -x to preserve directory structure

Prophyt,

I am not sure whether you found out the solution. Can you try the following and let me know if it helps.

find `/edi/aiprod/report` -mtime +30 -exec cp -b --parents {} /edilogs/aiprodreport/ \;
ERRCODE=`echo $?`
if [ $ERRCODE != 0 ] ; then
echo "Failed" | mailx -s "EDI Report Maintenance" xxxxxx@xxxxx.com
exit
else
echo "Successful" | mailx -s "EDI Report Maintenance" xxxxxx@xxxxx.com
rm -r `find /edi/aiprod/report -mtime +30`
fi

Hey Tiger,

When typing in the command you provided I got the following error.

ksh: /edi/aiprod/report: cannot execute
Usage: find path-list [predicate-list]

Now as I said before I'm a beginner to UNIX as well as using commands. So I could be entering it in wrong. But I entered in the command exactly how it was from what you provided.

If your find brings the files you want, leave the cp out and pipe it into cpio:

Code:

find .... | cpio -dumpv /somenewdir/for/backup

It will create the directory structure based on the found files too.

Zaxxon thanks for the reply this worked. But is there a way to only copy the directory structure for just the directory containing the files I want to copy?

(EX. Using the method you gave)
/edi/aiprod/reports/DSSI/filename
/edi/aiprod/reports/MICHELIN/filename
/edi/aiprod/reports/CATERPILLAR/filename

/edilogs/aiprodreport/edi/aiprod/reports/DSSI/filename
/edilogs/aiprodreport/edi/aiprod/reports/MICHELIN/filename
/edilogs/aiprodreport/edi/aiprod/reports/CATERPILLAR/filename

Basically I would like to remove /edi/aiprod/reports and have it copied so /edilogs/aiprodreport/DSSI/filename
/edilogs/aiprodreport/MICHELIN/filename
/edilogs/aiprodreport/CATERPILLAR/filename
would be the end result of the copy. Hopefully that makes sense? I do EDI (Electronic Data Interchange) mapping and the software I use I specificy the output directory. Was thinking the simpilar I make the directory strucutre the easier it would be on me. But again thak you for your suggestion because it work good.