prometheus-api.client — eigenbahn/prometheus-api-client 1.0.1

Client library to the Prometheus HTTP API.

All methods take a connection conn as a first argument. It's just a simple map in the form:

{:url "<prometheus_base_url>"}

The :url field value must contain the base URL of the Prometheus instance. It is not expected to contain the "/api/v1" part.

E.g. to connect to a local Prometheus instance:

{:url "http://localhost:9090"}

alertmanagers clj

(alertmanagers conn)

Get an overview of the current state of the Prometheus alertmanager discovery.

This corresponds to the /alertmanagers endpoint.

alerts clj

(alerts conn)

Get the list of active alerts.

This corresponds to the /alerts endpoint.

api-prefix clj

Constant part in the API url. Can be overridden e.g. in case of passing through a reverse proxy.

build-information clj

(build-information conn)

Get various build information about the instance.

This corresponds to the /status/buildinfo endpoint.

clean-tombstones clj

(clean-tombstones conn)

creates a snapshot of all current data into snapshots/<datetime>-<rand> under the TSDB's data directory. The created directory is returned as a response.

If fails, returns an exception (Slingshot Stone to be precise) with :status 204.

This corresponds to the /admin/tsdb/clean_tombstones endpoint.

config clj

(config conn)

Get the currently loaded configuration file.

This corresponds to the /status/config endpoint.

content-level clj

Level of content returned for API response calls. Valid values:

• ::http-client: raw response from clj-http.client, good for debugging
• ::body: HTTP body parsing into a clojure data structure
• ::data: "data" part of the prometheus response
• ::best: only the most sensible data for each endpoint (default)

convert-result clj

If true, parse convert the results into more clojuresque data structures. Time series are converted into maps, their timestamps converted into inst and scalar values parsed.

Default value is true.

Gets ignored if content-level is ::http-client

delete-series clj

(delete-series conn series-matchers & {:keys [start end]})

Deletes data for a selection of series (matched by a list of series-matchers) in optional time range (bound from :start to :end).

If fails, returns an exception (Slingshot Stone to be precise) with :status 204.

The actual data will still exist on disk and will be cleaned up in future compactions. This can be forced by calling the clean-tombstones API endpoint.

This corresponds to the /admin/tsdb/delete_series endpoint.

flags clj

(flags conn)

Get the instance configuration flags.

This corresponds to the /status/flags endpoint.

labels clj

(labels conn & {:keys [start end]})

Get the list of label names.

Optionally, we can filter on those present in a time range (with :start and :end).

This corresponds to the /labels endpoint.

metadata clj

(metadata conn & {:keys [metric limit]})

Get metadata about metrics.

Same as targets-metadata except doesn't contextualize with target names.

Optionally, we can filter on a specific :metric. A :limit can also be optionally specified.

This corresponds to the /metadata endpoint.

query clj

(query conn q & {:keys [at timeout]})

Query data with q at an instant :at. When :at is omitted, last data point is returned. Optionally, a :timeout can be specified.

If you cant to query over a time range, use query-range.

This corresponds to the /query endpoint.

query-range clj

(query-range conn q start end step & {:keys [timeout]})

Query data with q over a time range (from start to end with step duration between points). Optionally, a :timeout can be specified.

If you cant to query only at a given instant, use query.

This corresponds to the /query_range endpoint.

rules clj

(rules conn & {:keys [type]})

Get the list of alerting and / or recording rules.

Optionally, we can filter on the :type of rule. Valid values are: "alert" & "record"

This corresponds to the /rules endpoint.

series clj

(series conn series-matchers & {:keys [start end]})

Query series according to list of series-matchers.

Optionally, we can filter on those present in a time range (with :start and :end).

This corresponds to the /series endpoint.

snapshot clj

(snapshot conn & {:keys [skip-head]})

Creates a snapshot of all current data into snapshots/<datetime>-<rand> under the TSDB's data directory. The created directory is returned as a response.

Setting :skip_head to true allows skipping snapshotting data present only in the head block (not yet compacted on disk).

This corresponds to the /admin/tsdb/snapshot endpoint.

targets clj

(targets conn & {:keys [state]})

Get the list of targets gathered through target discovery.

Optionally, we can filter on the :state of the targets. Valid values are: "active", "dropped" & "any" (same as when left empty)

This corresponds to the /targets endpoint.

targets-metadata clj

(targets-metadata conn & {:keys [label-matcher metric limit]})

Get metadata about metrics scrapped from targets.

Optionally, we can filter on a specific :metric and :label-matcher ("match_target" URL parameter). A :limit can also be optionally specified.

This corresponds to the /targets/metadata endpoint.

tsdb-stats clj

(tsdb-stats conn)

Get various various stats about the cardinality of the TSDB.

This corresponds to the /status/tsdb endpoint.

values-for-label clj

(values-for-label conn label & {:keys [start end]})

Get the list of values for a label name.

Optionally, we can filter on those present in a time range (with :start and :end).

This corresponds to the /label/<label_name>/values endpoint.