implement handling of systemd active (exited) status

[sigurdblueface]

Host operating system: output of uname -a

uname -a
Linux test-zone-1 3.16.0-4-amd64 #1 SMP Debian 3.16.51-3 (2017-12-13) x86_64 GNU/Linux

node_exporter version: output of node_exporter --version

node_exporter, version 0.17.0 (branch: HEAD, revision: f6f6194)
build user: root@322511e06ced
build date: 20181130-15:51:33
go version: go1.11.2

node_exporter command line flags

  - '--path.rootfs=/'
  - '--collector.systemd'

Are you running node_exporter in Docker?

yes. via docker compose
docker -v
Docker version 18.06.3-ce, build d7080c1
docker-compose -v
docker-compose version 1.24.0-rc1, build 0f3d4dda

What did you do that produced an error?

suddenly my service became active (exited)

What did you expect to see?

I expected to see a change of a metric value
node_systemd_unit_state{state=active}

What did you see instead?

metric stood at value of 1.

and that there's no difference between active (running) and active (exited) statuses for the exporter.

Is there a possibility to handle such cases?

[SuperQ]

Hrm, good question, sometimes "active exited" is what people want. I guess we should have a separate state bit for this.

[discordianfish]

Wondering (once again) whether we should just spin out all the systemd stuff from the node-exporter..

[SuperQ]

@discordianfish There is a systemd_exporter being worked on, should replace the node_exporter stuff eventually.

[EpiqSty]

btw, seems systemd_exporter also not going to support unit substates before facing the real demand for such advanced feature
¯\_(ツ)_/¯
prometheus-community/systemd_exporter#12