Mike was looking into this earlier
http://jira.observium.org/browse/OBS-2043?filter=-2
However snmpv2 never had the snmp context implementation, so you would have to map the community string to a specific context on the device itself (thus vrf). Some nms solve this by adding the main device and then adding an additional list of communities to poll for vrf specific stuff (ip, bgp, etc)
SNMPv3 to my understanding has this implemented where it can switch between context using the same authentication; so a single entry would suffice, however I’m not sure it enumerates all the different contexts it can switch to in the ‘default’ context. And even so, on the devices we manage, contexts have to be defined manually on the device itself anyway :< even if in 99% of the cases they map one to one on a vrf.
Also; feedback from a Cisco employee regarding this https://supportforums.cisco.com/t5/xr-os-and-platforms/monitor-bgp-sessions-...
Anyhow, it’s a feature I believe many of us are waiting for. Some people reached out after sending it on the mailing list before.
Kind regards
From: observium [mailto:observium-bounces@observium.org] On Behalf Of Adam Armstrong Sent: donderdag 24 mei 2018 22:32 To: Observium observium@observium.org Subject: Re: [Observium] BGP polling inside VRF
Hi,
I looked into this recently for someone else, and it doesn't seem possible for an NMS to externally learn what contexts to poll to reach the VRF data.
It seems to require everything to be configured manually on both the router *and* the NMS, which is, tbqh, totally insane.
I don't understand what was so difficult about just adding another index and/or OID to the existing MIBs. The context stuff is nearly useless due to its complexity and manual configuration overhead, as evidenced by no one using it.
There may be vendors which provide a table that gives us pollable context data, but I have not yet found one :)
Adam.
------ Original Message ------
From: "Jan Horacek" <tarrendine@gmail.com mailto:tarrendine@gmail.com >
To: "Observium" <observium@observium.org mailto:observium@observium.org >
Sent: 2018-05-24 21:27:14
Subject: Re: [Observium] BGP polling inside VRF
Hi Adam,
this is not the answer I wanted hear, but understandable.
Thanks for quick reply
Jan.
2018-05-24 22:17 GMT+02:00 Adam Armstrong <adama@memetic.org mailto:adama@memetic.org >:
Hi,
This is not currently possible (and is pretty difficult to make possible).
The decision to do VRF stuff inside contexts was a stupid one on the part of the vendors, since it's not easily possible to know what contexts exist as an NMS, so everything has to be manually configured. This, of course, is plain absurd.
adam.
------ Original Message ------
From: "Jan Horacek" <tarrendine@gmail.com mailto:tarrendine@gmail.com >
To: "Observium" <observium@observium.org mailto:observium@observium.org >
Sent: 2018-05-24 20:57:03
Subject: [Observium] BGP polling inside VRF
Hi guys,
I`m not sure if this question was already answered or not, but I was asked to do research how Observium could monitor BGP peers inside the VRFs to alert us if any peer flaps or is down.
I can set up alert notification based on the syslog, but this module is sending emails only and doesn`t show notification on the main dashboard.
Do you have any idea if would be possible support snmp-context for every BGP peers in observium? Using SNMPv3 on all routers or how to display alerts on the dashboard based on syslog policy?
Thanks
Jan
_______________________________________________ observium mailing list observium@observium.org mailto:observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium https://emea01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fpostman.memetic.org%2Fcgi-bin%2Fmailman%2Flistinfo%2Fobservium&data=02%7C01%7CStef.Renders%40arxus.eu%7C506732d935ae4d083de508d5c1b59a19%7C49c3d703357947bfa8887c913fbdced9%7C0%7C0%7C636627908078917412&sdata=ARpRPl98DYlb5csPzxvdLkAIFygQ25m3Hmei3orWRB8%3D&reserved=0