Monitoring plugin check_rancher2 1.5.0 released: jq is now required instead of jshon, fixes cluster health check

Written by Claudio Kuenzler - 0 comments

Published on April 14th 2021 - Listed in Docker Kubernetes Rancher Internet Network Cloud Monitoring

---

A new version of check_rancher2, a monitoring plugin to monitor Kubernetes clusters managed by Rancher 2, is available.

With the release 1.5.0, there is a potential breaking change, as the plugin now requires the jq command instead of jshon to parse the json data from Rancher's API. However most distributions should allow an easy install of jq, as this command should be available in the standard repositories (better availability than jshon).

Debian/Ubuntu:

ck@debian:~$ sudo apt-get install jq

RHEL/CentOS (using EPEL repositories):

root@centos:~$ yum install jq

SLES/openSUSE:

root@suse:~$ zypper in jq

The reason for this is that jq allows to use json filtering which works much like a search. This was used to fix a problem where check_rancher2 falsely informed about an unhealthy cluster:

$ ./check_rancher2.sh -H rancher2.example.com -U token-xxxxx -P "secret" -S -t cluster
CHECK_RANCHER2 CRITICAL - controller-manager in cluster "local" is not healthy -|'clusters_total'=5;;;; 'clusters_errors'=1;;;;

The information is actually wrong, the "local" cluster at that point was absolutely OK. However another cluster actually had a problem.

With the version 1.5.0 and the change to jq, this is now correctly filtered and the bug is fixed. The affected cluster is now shown correctly in the cluster health check:

$ ./check_rancher2.sh -H rancher2.example.com -U token-xxxxx -P "secret" -S -t cluster
CHECK_RANCHER2 CRITICAL - "controller-manager" in cluster c-tx5fl is not healthy - "scheduler" in cluster c-tx5fl is not healthy -|'clusters_total'=5;;;; 'clusters_errors'=1;;;;

Documentation of check_rancher2 was updated, too.

Add a comment

Show form to leave a comment

Comments (newest first)

No comments yet.