NAV 3.13.0
Eventengine logs a lot of topology problems of two kinds.
One regarding the NAV-server itself:
2013-02-18 11:23:37,173 [WARNING nav.eventengine.topology] NAVServer('129.242.5.182') topology problem: router ma-gsw3.infra is up, but not in VLAN graph for <Prefix: 129.242.5.0/24 (vlan 103 (nett,srv,5-nettet))>. Defaulting to 'reachable' status.
The other kind regarding some (but not all) random devices:
2013-02-18 11:23:37,740 [WARNING nav.eventengine.topology] nfh-b100-nc.prn topology problem: router nfh-gsw.infra is up, but not in VLAN graph for <Prefix: 10.253.19.0/24 (vlan 600 (uit,prn))>. Defaulting to 'reachable' status.
Is there anything we can do to in our setup to eliminate the warnings (except changing the debug level...)?
--Ingeborg
On Mon, 18 Feb 2013 11:31:08 +0100 Ingeborg Hellemo ingeborg.hellemo@uit.no wrote:
NAV 3.13.0
Eventengine logs a lot of topology problems of two kinds.
One regarding the NAV-server itself:
2013-02-18 11:23:37,173 [WARNING nav.eventengine.topology] NAVServer('129.242.5.182') topology problem: router ma-gsw3.infra is up, but not in VLAN graph for <Prefix: 129.242.5.0/24 (vlan 103 (nett,srv,5-nettet))>. Defaulting to 'reachable' status.
The other kind regarding some (but not all) random devices:
2013-02-18 11:23:37,740 [WARNING nav.eventengine.topology] nfh-b100-nc.prn topology problem: router nfh-gsw.infra is up, but not in VLAN graph for <Prefix: 10.253.19.0/24 (vlan 600 (uit,prn))>. Defaulting to 'reachable' status.
Is there anything we can do to in our setup to eliminate the warnings (except changing the debug level...)?
Yes/no/maybe.
The WARNING level may be too high for these messages in the first place. They only appear when things stop responding and the event engine cannot make sense of the detected VLAN topology. I'll take your opinion on this any day.
The message tells you (or me, at least) that your NAV server is 129.242.5.182. The event engine thinks this is on VLAN 103, prefix 129.241.5.0/24 and that the default router for this VLAN is ma-gsw3. When interpreting NAV's topology data, however, it was, for some reason, unable to find the ma-gsw3 router in VLAN 103 topology.
The second message tells more or less the same story about the nfh-b100-nc device on VLAN 600 and the nfh-gsw router.
It's a shame you didn't participate in the beta testing period, Ingeborg :) If these problems are reproducible on tromso-vk, I could take a closer look at the data from here.
morten.brekkevold@uninett.no said:
When interpreting NAV's topology data, however, it was, for some reason, unable to find the ma-gsw3 router in VLAN 103 topology.
And why is that? Where can I find out what NAVs topology looks like? Because ma-gsw3 should indeed be a part of the VLAN 103 topology.
morten.brekkevold@uninett.no said:
If these problems are reproducible on tromso-vk, I could take a closer look at the data from here.
You will find similar errors on tromso-vk (different hosts but same type of error).
--Ingeborg
On Tue, 19 Feb 2013 13:48:46 +0100 Ingeborg Hellemo ingeborg.hellemo@uit.no wrote:
morten.brekkevold@uninett.no said:
When interpreting NAV's topology data, however, it was, for some reason, unable to find the ma-gsw3 router in VLAN 103 topology.
And why is that? Where can I find out what NAVs topology looks like? Because ma-gsw3 should indeed be a part of the VLAN 103 topology.
I can't say why just from this information alone. It just means that event engine expected the router to be in the in the graph built for the VLAN 103 topology, since it determined that this was the router for that VLAN, yet it didn't find it in the graph when searching for a path.
The error message is there just for this purpose: So that we are made aware of possibly fishy topology detection and can debug the data/algorithm. In your case, we are exceptionally privileged in that we have access to debug directly on your test server :)
morten.brekkevold@uninett.no said:
If these problems are reproducible on tromso-vk, I could take a closer look at the data from here.
You will find similar errors on tromso-vk (different hosts but same type of error).
If you could drop med an e-mail with the names of some to look at, compared with information from the real world, I'll take a look.
-
Ingeborg Hellemo -
Morten Brekkevold