pleskrestore disk space error - URGENT

[kieranjones]

I am attempting to restore a backup made with the plesk backup command line utilities, it's made the backup ok and is even able to restore it, but the problem is this.

It didn't seem to be able to restore databases. Websites have been restored without their respective MySQL tables and things. Why is this?

I decided to remove the offending domains and restore those from the backup again individually but I now get an error informing me that /dev/hda1 has run out of space. Understandable as when I check to see how much space /dev/hda1 has, it's only 1GB. It seems that the pleskrestore function unpacks the whole backup file to /root/restore which means that /dev/hda1 quickly runs out of space as /root is located within /dev/hda1

This probably explains the problems I had restoring the entire server backup because it was running out of space. The question is, how do I get around this? Is there some way to tell the pleskrestore tool to use a different directory to unpack into first?

This is really really urgent to do as many of the websites I host are currently offline.

Thank you for your help.

On a side note, I HATE 1&1 with a vengence now. This is all their fault.

 

[blank_page]

Hi,

Why don't you try to link that directory to a bigger partition.

 

[kieranjones]

How do I do that? I'm confident with computers and am happy to do anything at all via command line, but just don't know how. I shall Google it and see what I can work out

 

[kieranjones]

Hmm... I can't work it out, found lots of results for Macs but not for Linux... any ideas?

 

[JDwebsites]

You need to do a symbolic link for the root/restore directory. Here is an example.

if there is already a root/restore directory do this:

mv /root/restore /home/restore
ln -s /home/restore /root/restore

If no root/restore directory do this:

mkdir /home/restore
ln -s /home/restore /root/restore

Dean

 

[kieranjones]

Ah ha! That worked perfectly! Thanks JDWebsites. The pleskrestore tool reports errors when using a linked directory as it tries to delete the directory and finds it can't, but that isn't a problem as I don't want it to delete anyway.

 

[sonshine]

We have a similar problem.

I remember in plesk 7 it was no problem to restore a single domain. That was quick an harmles.
Now in plesk8 it takes hours for just one very small domain on a server with hundreds of domains (the full backup has a size of 24 GB).

We get a full partition very soon and have to stop the process to not influence the running server.

Why have restores to perform that bad since version 8?
Can someone explain that and what we could do against it?

BTW: After creation of the map file, I removed all other domains I am not interested in there (in our second try). But I think this is irrelevant (and not harmful) I guess?