<[EMAIL PROTECTED]> wrote:
} There are several vendors now offering "failsafe" capabilities. For
}instance, having a dual ported dual redundant RAID controller, so that
}two systems can access (different) RAID sets on the same set of RAID
}controllers, and if one server fails, the RAID sets can be mounted on
}the other server and ip addresses get pushed around, etc. etc. and
}life goes on with users realizing a system has failed. (Well,
}that's the theory at least).
} o mount the RAIDs from the failed machine on the good machine
} o restart bosserver and fs processes
} o run vos syncvldb <good server> <just mounted partitions>
} o may need to run vos changeaddr <failed ip> <good server ip>
We've not done this with AFS, but we're using DEC's DECsafe
(aka Available Server Environment) for an Oracle server.
It works pretty much as you've outline above. When a "service"
moves from one machine to another:
The disks assoicated with a service are mounted on the new machine
In addition to the IP address each machine has, there is an
IP address for the service which is moved to the new machine
(ifconfig ... alias)
Then a "start script" associated with the service is run.
If the service was cleanly moved from the old machine (e.g.,
a planned service migration rather than a machine fail), then
the opposite sort of things happened on it first (stop script,
loss of ip address, unmount).
It would seem that you could fit an AFS fileserver into this
without too much work.
John
--
John Hascall, Software Engr. Shut up, be happy. The conveniences you
ISU Computation Center demanded are now mandatory. -Jello Biafra
[EMAIL PROTECTED]
http://www.cc.iastate.edu/staff/systems/john/welcome.html <-- the usual crud
