(08:38:00) tanders: rstory: what is/was the reason for the "registering pdu failed: 263!" error for subagents in 5.2.2?
(08:53:15) rs: duplicate registration of the .1, .2 and .3 root oids
(09:28:37) tanders: is there a way around this? I don't like to see this on every startup
(09:51:43) dts12 [n=dts12@pub-nat.pclab.csc.liv.ac.uk] entered the room.
(09:56:32) hardaker [n=wes@dcn236-43.dcn.davis.ca.us] entered the room.
(09:58:26) rs: tanders: i dunno.. i don't like to muck about with the registration code..
(09:59:39) rs: you could hard code a check right before the registration is sent to the master, and supress it iff oidlen == 1 and context == NULL
(10:00:53) rs: although, in theory, the master might not support *any* mibs...
(10:02:57) crash3m left the room ("Leaving").
(10:03:40) rs: wow, dts12 dropped in w/out anyone even talking about him.. :-P
(10:04:09) dts12/dave: I don't think it matters whether the master supports any MIBs or not.  The root registrations don't provide any information, so shouldn't propogate through to the master agent.
(10:04:28) dts12/dave: Oh, and I know you would, so I though I'd pre-empt you :-)
(10:04:31) rs: what if the sub wants *all* requests..
(10:05:10) dts12/dave: Hmmmm.... that's a slightly different situation from the framework registrations.
(10:05:21) dts12/dave: Perhaps register those at a lower priority?
(10:05:40) dts12/dave: Or register them *before* setting up the AgentX link?  (Don't know if that's feasible)
(10:06:32) rs: i wonder what would happen if they weren't registered at all for a subagent...
(10:08:56) dts12/dave: Report back when you've found out :-)
(10:09:06) rojer is now known as rojer|way
(10:09:40) rs: heh.. don't hold your breath!