[OpenAFS] Understanding questions backup volume

Rainer Toebbicke rtb@pclella.cern.ch
Thu, 09 Feb 2006 13:35:21 +0100

Frank Burkhardt wrote:

>>At least a RO copy could be converted to a RW volume in nearly NO time,
>>but a backup volume?
> A backup volume can't because it's stored as a diff against the RW which means,
> it's damaged when the RW is.

As it stands this is incorrect: at the time of the clone a backup 
volume *is* the R/W volume (in the sense that its vnode files point to 
exactly the R/W's vnodes). It is when files on the volume change 
*they* are copied to new vnodes (here multi-GB file begin to become 
cumbersome). In that sense the R/W is stored as a diff to the backup, 
but with several clones you'd call this something else.

Conceivably however you could have a healthy backup volume and all 
files in the R/W somehow damaged (those modified since the clone). In 
that case it might make sense to "convert" the backup to a R/W. 'vos 
dump|vos restore' should do if you're prepared to pull out the hammer.

Rainer Toebbicke
European Laboratory for Particle Physics(CERN) - Geneva, Switzerland
Phone: +41 22 767 8985       Fax: +41 22 767 7155