Jan Wieck wrote:
>> Creating invalid nodes strikes me as the wrong way to do things.
> 
> Chris and I talked about it and I didn't like the idea at all. 

I'd have appreciated if your private discussion would have been a bit
more public, as I triggered it and have been waiting for any answer for
quite a while.

> The thing 
> is, all those "virtual" nodes would represent are admin stations, no? 

Yes. BTW, if a machine is a real node also, it can reuse that node's
path information happily.

> And all those are good for is to store their admin conninfo as sl_path 
> entries. Considering the amount of extra care to be take 

All I could locate was preventing to create listens for those nodes.

> (so that this 
> dead info doesn't affect any failover/switchover ... don't we already 
> have problems enough there) compared to storing that data in a separate 
> table that is automatically propagated ... I suggest separate table.

No problem with that (would need two tables, sl_adminnode and
sl_adminpath). Unless this goes into a 1.1.2, we'll have a gap of
administrability between 1.0 and 1.2.

I don't know failover/switchover handling in detail so far, that's why I
asked for feedback. Since a disabled node can't create events, and
without listen won't receive events (and thus won't hold up sl_event) I
didn't see problems in the first place (and am still wondering where
they should come from).

Regards,
Andreas