Shell Programming and Scripting

Ok. I am using one cluster to remotely write into another cluster. I can scp to copy files from cluste1 to cluster2, however, I need to copy a file that already exists in cluster 2 and replace it with the new file. I was thinking I could either remove the file in cluster 2 (which I don't know how to do) then use scp to copy the new file over, or use some variation of scp to overwrite the existing file in cluster2 with the new file in cluster1 (which I don't know how to do).

So lets say I do a

Code:

 scp directory/file1 user@host:directory/file1

but file2 in user@host already exists.

---------- Post updated at 10:39 PM ---------- Previous update was at 09:23 PM ----------

Ok. I'm trying to do this another way. If I can execute a script from cluster one to cluster 2 then I won't need to scp the file. I can seem to get it to work. This is what I have so far.

Code:

sh samplescript2.sh

samplescript2.sh has the following code

Code:

ssh user@host2 /bin/bash < /directory/samplescript.sh

samplescript.sh has the following code

Code:

cd /home/user/directory

yy=`date +%Y`
mm=`date +%m`
dd=`date +%d`

echo "Today is : $yy $mm $dd"

echo `pwd`

# Edit latest.txt file
  echo "${yy}${mm}${dd}00"            >latest.txt

When I execute the script I can see the date and the directory but nothing happens to the latest.txt file. Any ideas?

I can not reproduce that scp problem. scp overwrites any file that exists. I could also not find any option in it's man page to force overwriting or similar.

Maybe checkout this behaviour on other boxes of yours.

Ok. It worked. I simply made a mistake with my directory address in the script. Both options worked. Thanks.