[OpenAFS] Number of volume limits

Jeffrey Hutzelman jhutz@cmu.edu
Mon, 20 Aug 2001 15:51:07 -0400 (EDT)


On Mon, 20 Aug 2001 Craig_Everhart@transarc.com wrote:

> Excerpts from mail: 13-Aug-01 Re: [OpenAFS] Number of vol.. Jeffrey
> Hutzelman@cmu.ed (1133*)
> 
> > > Which is the limit on the number of volume created in one cell ?
> > > Is there a limits on the number of volume by vicep ?
> 
> > There is no practical limit on the number of volumes in one cell.  The
> > volume ID counter will run out at 4294967295, but to reach that, you'd
> > have to create one volume per second, every second, for nearly 120 years.
> 
> There's perhaps no practical limit, but the lower limit is probably
> imposed by the size of the VLDB entries in the VLDB, knowing that the
> VLDB itself is limited to 2G bytes.

Hm; good point.  I hadn't considered the impact of larger-than-2GB
databases, which of course aren't supported.  Somehow, I don't think this
is actually going to be a problem for anyone anytime soon; anyone who has
enough volumes for the VLDB to be >2GB is (IMHO) also pushing the safety
margin on the volume ID space. 

-- Jeff