Checks check docker swarm nodes

#CHECK_DOCKER_SWARM_NODES

Script: check_docker_swarm_nodes

check_docker_swarm_nodes shows the node state of a docker swarm cluster. It checks the nodes are ‘active’ and ‘ready’ (not in drain state).

This check sends performance data.

#Requirements

• docker Docker must be installed
• jq must be installed - commandline JSON processor
• sudo permissions on docker command

icingaclient ALL=(ALL) NOPASSWD:SETENV: /usr/bin/docker

#Standalone installation

From this repository you need next to this script:

• inc_pluginfunctions shared function for all IML checks written in bash
• inc_dockerfunctions.sh

#Syntax

______________________________________________________________________

CHECK_DOCKER_SWARM_NODES
v0.1

(c) Institute for Medical Education - University of Bern
Licence: GNU GPL 3

https://os-docs.iml.unibe.ch/icinga-checks/Checks/check_docker_swarm_nodes.html
______________________________________________________________________

Show docker swarm nodes

It returns 
  CRITICAL if a node is not 'ready' or availability is not 'active'
  OK if all nodes are up and running
  UNKNOWM if
    - docker or jq were not found
    - docker data were not fetched
    - docker cannot be connected
    - This host is not a swarm manager

USAGE:
  check_docker_swarm_nodes [OPTIONS]

OPTIONS:

  General:
  -h, --help        this help
  -d, --debug       Debug; Flag: add json of all docker nodes

  Connect to docker:
  -p, --path        Custom directory for docker binary
  -t, --target      Custom docker target; value for DOCKER_HOST
                    Needed only if Docker does not run on a unix socket or
                    multiple users run a rootless docker daemon.

  Flags:
  -n, --nodes       Flag: Show services
  -e, --errors      Flag: Show containers on error only

EXAMPLES:
  check_docker_swarm_nodes
    Show a short status of all nodes

  check_docker_swarm_nodes -e
    Show short status with all nodes. Non OK services will be shown.

  check_docker_swarm_nodes -s
    Show long status with all services and its state.

  check_docker_swarm_nodes -p /usr/bin
    Show status of all containers. The docker binary will be searched in
    given path first - then in all other dirs of $PATH

  check_docker_swarm_nodes -d
    Add json with all docker node infos

#Parameters

None.

#Examples

#Short status

$ ./check_docker_swarm_nodes

OK: Docker swarm services 3 ... OK: 3  errors: 0
 
 |nodes-ok=3;;;0;3 nodes-critical=0;;;0;3

#Long status

Without given options it returns a short info. To show all nodes of a swarm cluster use -nor --nodes:

$ ./check_docker_swarm_services -n

OK: Docker swarm services 3 ... OK: 3  errors: 0
 [OK]       f89v7jn0uxvnla9agi6fmw2p7 *   swarm-0.example.com   Ready     Active         Leader           29.8.1 
 [OK]       vqd6g506wjl00zipfo9q2ed3h     swarm-1.example.com   Ready     Active         Reachable        29.8.1 
 [OK]       wb72ldwesr8ghz7hf5q6os8ud     swarm-2.example.com   Ready     Active                          29.8.1 
 
 |nodes-ok=3;;;0;3 nodes-critical=0;;;0;3 

#Troubleshooting

#Missing environment

sudo: sorry, you are not allowed to preserve the environment
UNKNOWN: No data. Unable to fetch Docker information.

Solution: you did not set SETENV in the sudoers config file

icingaclient ALL=(ALL) NOPASSWD:SETENV: /usr/bin/docker

#Docker not found

If you get an error message that docker was not found then it is not reachable in ofe of the directories of $PATH. This can happen especialy in rootless docker instances.

Solution: Use the parameter -p /usr/bin to define a target. Copy the docker binary there.