Failover Questions

[David W. Hankins]

On Tue, Jan 23, 2007 at 09:59:47PM +1100, Glenn Satchell wrote:
> I vaguely remember David Hankins posting some time ago that the actual
> key used to determine the failover peer was the IP addresses and ports.
> And that this was going to change in a later release (maybe 3.1.0?) to
> actually use the peer name.

That's correct, in 3.1.0 we use the peer name to find the failover
session rather than the TCP socket's remote address.  The failover
draft requires the name be used, so this was one of the changes
while we were trying to sync up to -12.

Note that there's also an MD5 digest authentication mechanism in
failover (which we don't yet implement) which makes this look
more like a 'username'.  It might be easier to think of it in
that context.

It's kind of unfortunate that 'failover peer "foo"' reads like the
name of the peer rather than the name of the failover 'relationship'.

-- 
David W. Hankins	"If you don't do it right the first time,
Software Engineer		you'll just have to do it again."
Internet Systems Consortium, Inc.	-- Jack T. Hankins