[OpenAFS] Fwd: IBM -> OpenAFS solaris 7 upgrade?

Hartmut Reuter reuter@rzg.mpg.de
Tue, 03 Aug 2004 16:32:58 +0200


If you upgrade your fileservers this way, be sure the OpenAFS fileserver 
is not
built with "--enable-namei-fileserver". Otherwise you have to start with a
new empty fileserver and to move all your volumes from the old to the 
new one.

Hartmut

John W. Sopko Jr. wrote:
> Hey everyone,
> 
> I would appreciate input on my following situation:
> 
> I am currently running 3 Solaris 7 Transarc/IBM database/file servers.
> They are running Transarcs patch 8, (Base configuration afs3.6 2.45
> with the patch to fix that ubik date problem). Transarc/IBM has stopped
> supporting Solaris 7 and we do not want to pay for support to get
> updates anyway.
> 
> We have been running OpenAFS file servers and clients on Red Hat
> linux and it seems to work well. I have tested the OpenAFS file
> server only on a Solaris 7 machine. I would like to update my 3
> Solaris 7 boxes to OpenAFS. Can I just drop in the binaries like
> I normally would for an AFS patch upgrade? I would:
> 
> Put the /usr/afs/bin binaries in place on all 3 db/file servers and
> restart the lowest to highest IP address. Also put the special fsck
> in place.
> 
> I guess some of the questions are:
> 
> - Is OpenAFS 100% compatible with all my current db servers databases?
>   (including the bdb backup database)
> 
> - When I upgrade should I stop all the db/file servers then bring them
>   up one at a time so I am not momentarily running the Transarc/IBM
>   versions at the same time as the OpenAFS versions?
> 
> - Anything else to watch out for?
> 
> I plan on testing this on a test cell. First wanted to see if this
> is a legitimate upgrade path.
> 
> 


-- 
-----------------------------------------------------------------
Hartmut Reuter                           e-mail reuter@rzg.mpg.de
					   phone +49-89-3299-1328
RZG (Rechenzentrum Garching)               fax   +49-89-3299-1301
Computing Center of the Max-Planck-Gesellschaft (MPG) and the
Institut fuer Plasmaphysik (IPP)
-----------------------------------------------------------------