We use a maximum of about 128GB per server and have 40,000 volumes total and
haven't had a problem until we upgraded one of the fileservers from AFS 3.4a
to 3.5. It has 29,000 volumes on it and used to take about 7 minutes to
shut down the fileserver process and about 20 minutes to start it up. When
we upgraded to 3.5, we encountered a bug where it times out while shutting
down the fileserver process after 30 minutes and then kills it. So it now
takes 30 minutes to shut down, 60 minutes to salvage and 20 minutes to start
up the fileserver process. Transarc has been able to duplicate this
problem, but hasn't been able to provide us with a fix yet.
On Fri, 12 May 2000, Touretsky, Gregory wrote:
> In Intel Israel we have now 256Gb per AFS server (thinking about
> 512Gb/server).
> Partition size is 24Gb.
> Total # of volumes ~45.000 (including RO volumes)
> Total diskspace is ~8Tb
> We'd be glad to share more information about AFS installations in similar
> configurations.
>
> Gregory Touretsky
> IDC Computing / Systems Engineering Group
> Unix Server Platforms
> [EMAIL PROTECTED]
> > (+) 972-4-865-6377, Fax: 04-865-5999
> iNET: 465-6377, M/S: IDC-1B
>
>
>
> -----Original Message-----
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, April 20, 2000 5:51 PM
> To: [EMAIL PROTECTED]
> Subject: afs file server sizes
>
>
> I am wondering what people run as "large" afs file servers nowadays. What
> is
> the comfort zone in running:
> - large gb per server (or gb per partition)
> - large number of volumes per server
>
> How is recovery time on these large servers that crash? Any problems with
> volume cloning, backups, etc?
>
>
> Mark Giuffrida
> CAEN Systems Group
> [EMAIL PROTECTED]
>
>
>
>
>
- Mike
--------------------------------------------------------------------------
Mike Burns UNIX Systems Group
[EMAIL PROTECTED] Center for Academic Computing
+1 814 863 5606 The Pennsylvania State University
