I've recently set up a headless server to back up a number of very large film & editing files.
I SSH into the server and - following some best practice guides I found around - I've tightened security so that the server auto logs me off after 5 minutes of inactivity.
This is now creating a bit of a headache as I try to sort and move those files.
Issuing mv file1 /new_directory/ can take +5 mins and so I'm often logged off half way through.
Where multiple files are involved I've come back to find that some files seem to have transferred and others (at a seemingly arbitrary cut off point... or roughly what I reckon a 5 min timeout might have allowed) have not.
So my questions are:
- When issuing a
mvcommand, if it is cancelled whilst incomplete is there any risk of data loss? Might I be safer usingcp? - Is there any way that I could set a
cpormvcommand to go and then be sure that it will continue even after I've logged off or my ssh window has closed - Is there a way to do other things whilst a
mvorcpis taking place (and thus refresh my countdown timer)
2 Answers
Don't log in as such. Instead, from your local machine, run
ssh user@server mv /path/to/source /path/to/destThat should let you get around the problem.
On a more general note, I do believe you are being a bit paranoid here. Setting the idle time to 5 minutes and then facing this kind of issue seems silly. Just set it to something longer and avoid the problem.
As for your 1st question, mv will only delete the source files if the copy succeeded. As explained in info mv (emphasis mine):
5It first uses some of the same code that's used by `cp -a' to copy the requested directories and files, then (assuming the copy succeeded) it removes the originals. If the copy fails, then the part that was copied to the destination partition is removed. If you were to copy three directories from one partition to another and the copy of the first directory succeeded, but the second didn't, the first would be left on the destination partition and the second and third would be left on the original partition.
It is highly recommended to do long-running admin tasks only with tmux or screen (or Xvnc if you need it graphically). This shields you from network disconnects and allows reconnect. You could also use mosh, which is robust against network outages.
If you do not care so much about continue the session after your command has terminated, you can also run it detached with nohup mv ... (optional ending with & to background it immediatelly). You then need to monitor its execution with ps. Some shells can detach a running program (disown or not kill them like zsh's setipt NO_HUP) or use a background starter like at or cron.
