Add virtual chassis port status alerting
Closed, ResolvedPublic

Description

ayounsi created this task.Aug 2 2018, 8:37 PM
ayounsi triaged this task as Normal priority.
Restricted Application added a project: Operations. · View Herald TranscriptAug 2 2018, 8:37 PM
Restricted Application added a subscriber: Aklapper. · View Herald Transcript

Putting the script here the time I send a Gerrit CR.
It uses snimpy and the required MIBs can be obtained on https://apps.juniper.net/mib-explorer/index.jsp

mib-jnx-smi.txt
mib-jnx-chassis.txt
mib-jnx-ex-smi.txt
mib-jnx-virtualchassis.txt

$ python3 check_vcp.py --host asw2-a-eqiad.mgmt.eqiad.wmnet --community <secret>
OK: UP: 20

I then delete one side of a VC-link:
asw2-a-eqiad> request virtual-chassis vc-port delete pic-slot 0 port 48 member 2

$ python3 check_vcp.py --host asw2-a-eqiad.mgmt.eqiad.wmnet --community <secret>
CRIT: Down: 1 Unknown: 0
FPC N/A port N/A down

A couple limitations on the data exposed via SNMP:

  • Doesn't alert if the status is "Absent" (cable unplugged on both sides)
  • Doesn't return the FPC nor port # (present in the MIB but not in the SNMP reply)
  • [Untested] Probably doesn't alert if the link is UP but the adjacency not established (Junos bug?)

Change 458850 had a related patch set uploaded (by Ayounsi; owner: Ayounsi):
[operations/puppet@production] Icinga: add check_vcp (part 1)

https://gerrit.wikimedia.org/r/458850

Change 458850 merged by Ayounsi:
[operations/puppet@production] Icinga: add check_vcp (part 1)

https://gerrit.wikimedia.org/r/458850

Change 476604 had a related patch set uploaded (by Ayounsi; owner: Ayounsi):
[operations/puppet@production] Icinga, assign check_vcp to all VC switches

https://gerrit.wikimedia.org/r/476604

Change 476604 merged by Ayounsi:
[operations/puppet@production] Icinga, assign check_vcp to all VC switches

https://gerrit.wikimedia.org/r/476604

Mentioned in SAL (#wikimedia-operations) [2018-11-29T20:01:22Z] <XioNoX> Apply Icinga:check_vcp to all VC switches - T201097