Scalix Forums

I had a hard-drive fail in my Scalix E-mail server, running Fedora Core 4. From everything I've read I'm lead to believe that the user data that is backed up with sxbackup (I don't have a recent backup..) should be in /var/opt/scalix/ but when I browse to that directory using a bootable OS CD and mounting the drives (In Knoppix 4) I don't see anything in that directory. Am I looking in the wrong place? I found some files in /opt/scalix/template that look like they are the folders I need, but I don't have access rights to them for some, and the ones I do have access don't have any data in them....

Thanks whoever looks at this!

The mailstore data are all in /var/opt/scalix. /opt/scalix/template is used to build an empty mailstore, you won't find anything useful there.

Is it possible the mailbox data is on a different HDD / partition and that with Knoppix it simply doesn't get mounted properly and is therefore empty?

Well, I had a hard-drive failure. I've got 3 sda devices listed through knoppix, but i can only access two of them.... One of them is small and I think is the boot sector (I'm a Linux newb so I'm not sure) and the two drives are supposed to be in Raid 1 configuration, so theoretically they should be the same drive.... Knoppix is just wierd. It could be my drive is toast, and the data should be there and is just gone

I've got a 2 month old backup. My godly Linux server wasn't supposed to crash So, I am the fool on that one. I've really decided that my drives are toast, just thought I'd give it one last try... From what I understand, the information store was in /var/opt/scalix. The guy who installed the server told me that he didn't install to a non-default location and it should be there (he told me originally that it should be in the /var/opt/scalix folder)

Thanks alot Valerion, you were a lot of help

If your scalix mailstore was on software raid you might find this useful:

Knoppix probably won't assemble any existing software raid devices for you. Assuming your root fs wasn't on software raid (eg it was /dev/sda1) then mount it (e.g. on /mnt/root) & try to re-assemble any raid devices:



Or you could use the backup you mentioned to get a copy of mdadm.conf.

If you know the configuration of your raid device you don't need the conffile, just give the info on the command line - check the mdadm manpage.

Even if one drive's dead the array should still assemble, albeit in degraded mode. If so you should mount the raid device read-only & copy the data to a backup device. You can then remount it rw & use mdadm to remove the dead drive from the array & add a suitably prepared new drive.

NB: be careful not to use the '--build' option as this will initialise the raid component devices, destroying any existing data (madam is kind enough to warn you about this if it notices a drive/partition was already part of a raid device).

Even if you're content to use your older backup I would take the opportunity (if time allows) to play around with this as you'll find the experience invaluable.

Hey, thanks for the help guys! I gave up on the information for now and I'm going to load in my stagnant backup I appreciate your insight though