pre-creating reserved leases

[David W. Hankins]

On Thu, Mar 26, 2009 at 05:14:05PM -0500, Foggi, Nicola wrote:
> Or maybe it's related to the "bind update on xxx from xx rejected:
> incoming update is less critical than outgoing update" message I'm
> seeing on my secondary, maybe who owns the lease is getting lost.

this is just "dueling update suppression," i would find it very
unlikely that this log message means anything meaningful.  the
problem is the current code finds it hard to suppress a rejected
binding update (there are only accepts and rejects, and all rejects
are treated like critical errors).

> Definitely some bugs in lease_mine_to_reallocate and it's handling of
> "RESERVED" leases, so hopefully my bug report helps, not sure if it's
> proper as I mentioned in the report :(

this behaviour follows the draft's text...we will have to evaluate if
the optimization (to re-issue a released reserved-lease) is a safe
one.  there are other safe optimizations we cannot currently partake
because we only keep the lease's current state, not its previously
negotiated state.

i think we would have to know that both servers acknowledged the
reserved bit status, which isn't presently known.  but it's worth
evaluating if we don't need historical state in this case.

-- 
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
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <https://lists.isc.org/pipermail/dhcp-users/attachments/20090326/e51609fc/attachment.bin>