Backup and restore of huge mailboxes

[nighthawk]

up using Ok guys

I ended up using IMAPSync to transfer all the email from one instance to the other.

Everything worked like a charm transferring 25Gb of email from 31 accounts.

I tried to create a backup with the 8.8.15 version using zmmailbox -z -m {user} GetRestURL etc... etc... but the file created has always useless folders like Inbox!1, Inbox!2 and so on.
Every folder has a maximum file count of 500 mails.

If this format is handled correctly for a restore, I perfectly live with this. But unfortunately the restore procedure recreates the same structure found in the tgz format so with all the fancies folder... and it is useless!!!

I'm not interested in a solution like the one offered by Zextras becouse I need to run the backup at specific time and then deliver it to the backup machines (one internal and the other encrypted on a cloud).

There are some very interesting project around the backup via IMAP but no one has the possibility to insert the mail administrator password to permits the backup of the user emails with specific passwords.

If ImapSYNC had a function like this (i.e. not only sync between IMAP servers but also crate an mbox file or various eml files) this will be overkill.... But it hasn't these feature... unfortunately.

Any suggestion will be very much appreciated.

Thank you

[phoenix]

Have you looked at any of the wiki article scripts to see if they do what you want?

https://wiki.zimbra.com/wiki/Backup_and ... e_Articles

[L. Mark Stone]

up using Ok guys

<snip>
I'm not interested in a solution like the one offered by Zextras becouse I need to run the backup at specific time and then deliver it to the backup machines (one internal and the other encrypted on a cloud).
<snip>

Any suggestion will be very much appreciated.

Thank you

ZeXtras, and Network Edition Backup NG, offer Real Time Backups. When enabled, the backups are continuous and logically consistent at all points in time.

That means you can rsync the backup directory, snapshot the disk, or whatever at whenever specific time you wish, and have a standalone backup from which you can do a quick, complete system restore.

And because the backups are continuous, periodic "catch up" rsyncs throughout the day are fast, since you are only copying the new data over from the source (and deleting removed/changed data on the target).

Unless I misunderstand your goals, that would seem to accomplish what you want: near zero RPO and a very short RTO (see the Admin Guide for specific Disaster Recovery instructions: https://zimbra.github.io/adminguide/lat ... ith-backup).

Hope that helps,
Mark