CouchDB Prometheus Exporter
CouchDB metrics exporter for Prometheus
The CouchDB metrics exporter requests the CouchDB stats from the /_stats
and /_active_tasks
endpoints and
exposes them for Prometheus consumption. You can optionally monitor detailed database stats like
disk and data size to monitor the storage overhead. The exporter can be configured via program parameters,
environment variables, and config file.
Build the binary
You can find pre-build releases for different platforms at our GitHub Releases.
If you prefer to build your own binary or in case you'd like to build from the current master
,
you'll have to get and install a recent version of Golang for your platform, first.
Then, you have to perform the following commands in the cloned repository:
export GO111MODULE=on # in a Windows shell, please replace `export` with `set`
go get github.com/gesellix/couchdb-prometheus-exporter
Those commands will install the binary in your local GOBIN
directory, usually something like
$HOME/go/bin
. Please ensure that the directory is in your system's PATH
. Then the following
should work:
couchdb-prometheus-exporter --help
Run the binary
You can get an overview over possible configuration options with their defaults in the help screen:
couchdb-prometheus-exporter --help
Configuration is possible via:
- environment variables (e.g.
COUCHDB_USERNAME=admin
) - command line parameters (e.g.
--couchdb.username admin
) - configuration file
The configuration file format is the "properties" file format, e.g. like this:
couchdb.username=admin
couchdb.password=a-secret
Run it as container
docker run --rm -p 9984:9984 gesellix/couchdb-prometheus-exporter --couchdb.uri=http://couchdb:5984 --logtostderr
Please note that host names like localhost
won't leave the container, so you have to use non-loopback
dns names or ip addresses when configuring the CouchDB URI.
Logging
The couchdb-exporter uses the glog library for logging.
With the default parameters everything will be logged to /tmp/
.
Use --logtostderr
to enable logging to stderr and --help
to see all options.
CouchDB 2+ clusters
For CouchDB 2.x, you should configure the exporter to fetch the stats from one node, to get
a complete cluster overview. In contrast to CouchDB 1.x you'll need to configure the admin
credentials, e.g. like this:
couchdb-prometheus-exporter --couchdb.uri=http://couchdb:5984 --couchdb.username=root --couchdb.password=a-secret
Database disk usage stats
If you need database disk usage stats, add a comma separated list of database names like this:
couchdb-prometheus-exporter --couchdb.uri=http://couchdb:5984 --databases=db-1,db-2 --couchdb.username=root --couchdb.password=a-secret
Or, if you want to get stats for every database, please use _all_dbs
as database name:
couchdb-prometheus-exporter --couchdb.uri=http://couchdb:5984 --databases=_all_dbs --couchdb.username=root --couchdb.password=a-secret
Monitoring CouchDB with Prometheus, Grafana and Docker
For a step-by-step guide, see Monitoring CouchDB with Prometheus, Grafana and Docker
A complete example stack including multiple CouchDB instances, exporters, Prometheus, Grafana, etc. is available at examples/grafana/
, and can be deployed locally:
cd examples/grafana
docker swarm init
docker stack deploy --compose-file docker-traefik-stack.yml example
Examples
The examples
directory in this repository contains ready-to-run examples for
Credits
Thanks go to the Prometheus team, which is very active and responsive!
I also have to admit that the couchdb-prometheus-exporter code is heavily inspired by
the other available exporters,
and that some ideas have just been copied from them.
Last but not least, this project wouldn't be possible without users submitting issues,
feature requests and adding code contributions.
Thanks a lot!
Metrics Overview
The file README_metrics.md gives you an overview on the currently exposed metrics.