[rdiff-backup-users] Sorry for polluting the list - To make amends, here's a few tips & tricks

[Øyvind Skaar]

Hi All

Sorry for polluting the list. Lesson learned: Do not hack together your 
own auto-reply in the mail-client :( .


In an effort to make amends, here's a few random tips and tricks. 
Nothing earth-shattering, but hopefully someone, somewhere, at some 
point, will find it helpful.

Generally, rdiff-backup works well and does not require a lot of 
intervention. The biggest problem ATM is the we restore seldom, so an 
error can go unnoticed for a long time unless we catch it. The error 
messages are often horrible.


Rdiff-backup is very flexible and kind of low-level, which is a good 
thing. But it also mean it's not very useful without some additional 
scripting.

- As a bare minimum, check to see if rdiff-backup generated a
  statistics file for the run.
  There should be a "rdiff-backup-data/session_statistics.*.data" file.
  The data in it should make sense =)


- SSH's "ServerAliveInterval" helps long-running ssh connections
  stay alive:

    echo -e "ServerAliveInterval 5\n" >> ~/.ssh/config
    chmod 600  ~/.ssh/config


- Use a ssh key with rsync-backup like this:
  --remote-schema "ssh -i /path/to/key -C %s "


- Use "--terminal-verbosity" :
  Sometimes manual rdiff-backup operations takes a long time. It's nice
  to get some feedback on what's going on.
  "--terminal-verbosity 6" seems lik a good tradeoff.


- "--test-server" doesn't always work.
  Mostly use "--list-increments" to test the repository.


- Diff can help you compare folders as well:
  diff -rq --exclude=".*" /home/os/ /tmp/restored_files/




ø


--
Øyvind Skaar
Systemkonsulent, Opoint AS
Akersgata 28A, 0158 Oslo
Epost: address@hidden - Mobil (+47) 48278480