Skip to content

tjhop/saltstack-highstate-metrics

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 

Repository files navigation

saltstack-highstate-metrics

Saltstack is a great configuration management system, but it can be difficult to assess salt's performance, since salt does not natively expose application metrics (in Prometheus format or otherwise).

There is a salt extension that is being worked on, but until that is ready something else is needed. There are some community exporters such as this one and this more recent one, which provide varying levels of metrics as well. Both of these exporters run long lived services to expose proper HTTP metrics for prometheus to scrape -- the first works by running a highstate on an application-configured schedule, and the latter working by connecting to the salt controller's event bus to observe events.

This project takes a different approach -- it's simply a wrapper script written in Bash to initiate and run a salt highstate on a minion and parse the JSON output to generate metrics. The metrics get written to a textfile for collection by the Node Exporter's Textfile Collector.

Example of the metrics that get generated:

root@debian:~# cat /var/lib/prometheus/node-exporter/saltstack-highstate-metrics.prom
# HELP salt_highstate_disabled Static metric indicating if highstates are disabled. 0 is enabled, 1 is disabled.
# TYPE salt_highstate_disabled gauge
salt_highstate_disabled 0

# HELP salt_highstate_run_total Counter of how many times highstates have been run, via /usr/local/bin/saltstack-highstate-metrics.sh.
# TYPE salt_highstate_run_total counter
salt_highstate_run_total 164

# HELP salt_highstate_states Gauge indicating how many states were executed in the last highstate.
# TYPE salt_highstate_states gauge
salt_highstate_states 280

# HELP salt_highstate_states_failed Gauge indicating how many states failed execution in the last highstate.
# TYPE salt_highstate_states_failed gauge
salt_highstate_states_failed 0

# HELP salt_highstate_last_highstate_duration_seconds How long the last highstate took to run, in seconds.
# TYPE salt_highstate_last_highstate_duration_seconds gauge
salt_highstate_last_highstate_duration_seconds 247.57721

# HELP salt_highstate_last_highstate_seconds Unix epoch timestamp of when the last highstate was run via /usr/local/bin/saltstack-highstate-metrics.sh, in seconds.
# TYPE salt_highstate_last_highstate_seconds gauge
salt_highstate_last_highstate_seconds 1670349882

The wrapper script can then be run via any mechanism that allows for scheduled jobs -- systemd timers, periodic nomad jobs, etc.

About

Bash wrapper to run a salt highstate and generate Prometheus metrics from the JSON output

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages