Shibboleth Developers

[Chad La Joie <>]

Scott Cantor wrote:
> > nodes is thus navigable)  So, this is why I suggest you place each nodes 
> > IP address in every node's config.  It's not like the list should change 
> > all that often.  (Sorry for the very long winded side track there.)
>
>
> Right, just wanted to be sure I didn't have to take it out of every peer
> list to safely take a box down.

Yep, it's safe.  One of the things I always liked about the JBoss 
replication (which is just their services using the same library I'm 
using here) is that you could do rolling app server upgrades.  JBoss 
tends to have short release schedules with decent things in each release 
so being able to roll the upgrades, without service loss, was nice.

> One thing you might note is that there's really no amount of stickiness you
> can use to make asynch replication work. It's two different boxes making the
> connection for a given session, so I don't see how you'd ever link them. It
> works for fail-over mode, but that's about it.

Right, and some people set things up like that where they have a single 
active node and some number of standby nodes.  But your right, I should 
spell that out in the documentation.  Thanks.
--
Chad La Joie             315Q St. Mary's Hall
Project Sentinel         202.687.0124