[533] in Info-AFS_Redistribution
Re: some questions about using AFS to share core OS files
daemon@ATHENA.MIT.EDU (Stripes (or Josh, if you must))
Thu Jan 16 14:27:19 1992
To: Lyle_Seaman@transarc.com
Cc: Info-AFS@transarc.com
In-Reply-To: Your message of "Wed, 15 Jan 92 18:42:58 EST."
Date: Thu, 16 Jan 92 12:40:09 -0500
From: "Stripes (or Josh, if you must)" <stripes@eng.umd.edu>
Lyle_Seaman@transarc.com said:
[...]
>In regard to the earlier comments about reliability and availability,
>it should be clear that by replicating binaries on multiple servers
>and distributing the servers throughout your network, you can achieve
>very high availability, and protect against network partition. (ie,
>the probability that all the servers will be unavailable at the same
>instant is approximately (1- mean server availability)**NSERVERS)
That assumes the only reason a server will be unavailable is because the
server itself is down.
Real networks have other failure modes. Bad taps can reduce network bandwidth
to almost nothing for a whole section of a cable. Bad routers or LANbridges
can cut off dozens of clients (or servers) from the net at large. Broadcast
storms can cloge networks (try ping'ing the broadcast address with 1K packets
from a box with a fast ethernet, like a 4/65; instant no-net). Nameserver
problems may turn a working network into a broken one (either by refusing
to identify hosts, giving wrong info, or just sucking all the bandwidth
when mis-configured).
Of corse since you normally have the users data on an AFS volume anyway, what
good does it do them to be able to login to a minimal system, but have no
files?
[...]
--
stripes@eng.umd.edu "Security for Unix is like
Josh_Osborne@Real_World,The Multitasking for MS-DOS"
"The dyslexic porgramer" - Kevin Lockwood
"NFS only behaves properly..." - jik@athena
"...when the computer is not drawing power." - rbj@uunet
