[OpenAFS] switch from memcache to diskcache

[Adrian Knoth]

On Thu, Oct 31, 2002 at 03:53:26PM -0500, Derek Atkins wrote:

> First, I don't understand -- if your machines are diskless then
> how does a "disk cache" save startup time?  You're not saving data
> to any disk -- you HAVE no disk.

Ok, we have no disk for data. There is a 20GB-SCSI-disk. We use it
mainly for caching (afsd).
 
> _then_ shutdown AFS from within the root ramdisk (the same way it was 
> started), 

I guess there is no easy way to do this, but we'll see.
 
> Note that AFS was never designed or meant to be a root file system, 

Oh, it works very well in this way. :) If I could add a wish: ACLs
for files, not only for directories. SCNR :)

> Also note that you -can- have "dataless" workstations that have disks

We have.

> Yes, it adds aboout $100 to the cost of your machines, but frankly it saves
> your network.  

We migrated from workstation-installation (every node ran SuSE from disk)
to the AFS-root-version (now using Debian), so the disks were present
right from the beginning.

> There is no way your network could deal with a restart avalanche of 1000 
> diskless machines!

I think so, too. Perhaps with a lot of replication-servers, independend
switches and a time-multiplexed boot it would be possible, but to say it
again: you don't recognize any speedproblems with this cached approach.
It feels very "local" :) 

-- 
mail: adi@thur.de  	http://adi.thur.de	PGP: v2-key via keyserver

Ich bin nicht arbeitslos! - Jeden Tag kommt eine neue Stellung dazu!