We are running several instances (for different clients) Open Source zimbra and one network edition zimbra.

We have a dedicated backup server than runs the backups (including other directories on those servers) as described in the open source back procedure (allbeit slightly modified).

All the zimbra versions are in the 5.0.XX version series.

The modified backup script is a bashscript run on a crontab job from the back server.

Basically the zimbra portion of the backups run a hot sync of zimbra using rsync as such:

rsync -avzHkh --delete --progress --inplace <root>@<remote_host>:/opt/zimbra /<path_to_local_backup_of_zimbra>/
then the backup scripts calls

ssh <root>@<remote_host> sudo -u zimbra /opt/zimbra/bin/zmcontrol stop
then the script cold syncs then we restart the zimbra using
ssh <root>@<remote_host> sudo -u zimbra /opt/zimbra/bin/zmcontrol start
This is where the problem is - sometimes zimbra doesn't always come back up and ssh'ing into the server and running zmcontrol restart as zimbra locally on the machine always fixes the problems.

There are two zimbra's that never restart the logger process and I have to log in and run:

zmswatchctl start
zmloggerctl start
for all the zimbra processes to be up and going again.

Is running ssh <root>@<remote_host> sudo -u zimbra /opt/zimbra/bin/zmcontrol start a problem and if so how can it be fixed. Should I load this as a bug against zimbra.