When does the error appear? - I suppose not when you run the command as you have not defined any object to recover yet. Could it just be that there is unsufficient fres space at the destination?

It is after I select all the directories I want recovered and in the middle of recovery.

By that I mean both before and after the particular files recover fine.

There is tons of space, nearly twice as much as the entire mount point on the original server, so there should not be an issue there.

Is this an issue with the protocol that recover uses? What does it use?

TCP & RPC as usual.

I do not see this as an issue as it should have appeared already during the backup.

I could also imagine that the size info in the file index is too large (for whatever reason).

You should see it when you run "nsrinfo -v -t savetime client".

Hello,

check if your relocation directory is mounted on a 'real' filesystem, not on /.

And if you restore under AIX, check if you grant big files in your filesystem.

Nope, these are the files I am having problems with:

ashley@/datatel/coll18/production/apphome # du -g WWW*

1.00    WWW.STATE.FVAR

1.00    WWW.STATE.LOCAL

1.00    WWW.STATE.TX

Have you big files > 2 GB ?

It is a mounted jfs2 file sytem created from a logical volume using smit, here is what it looks like with df -g command

Filesystem              GB blocks      Free      %Used    Iused      %Iused      Mounted on

/dev/Colleague_lv    325.00         237.90   27%        784645     2%           /datatel

Is your recovery directory empty ?

Busy files ?

Raw files ?

recovery directory empty = no

Have you tried to recover only those files in an empty directory ?

Would type of file really make a difference?

Don't know, but exploring all that I can !

recovery directory empty = no

busy files = no - from tape onto a server that is not yet in use

raw files = maybe, i do not know exactly

Would type of file really make a difference?