PromLens is a web-based PromQL query builder, analyzer, and visualizer.
PromLens was initially created as a commercial software by PromLabs and then donated to the open-source Prometheus project by PromLabs and Chronosphere.
Please file issues on the GitHub issue tracker.
There are multiple ways of installing PromLens:
You can find pre-built release binaries for PromLens on the GitHub releases page.
Docker images are available on Docker Hub and Quay.
As a minimal example, you should be able to run PromLens like this:
docker run -p 8080:8080 prom/promlens
To build PromLens from source, you need:
- Go version 1.17 or greater.
- NodeJS version 16 or greater.
- Note: With NodeJS v17+ you may experience an
error:0308010C:digital envelope routines::unsupported
error. This can be worked around by either exporting the following environment variableNODE_OPTIONS=--openssl-legacy-provider
(please be aware of security considerations) or downgrading to NodeJS v16 (source).
- Note: With NodeJS v17+ you may experience an
- npm version 7 or greater.
Note: This is a non-exhaustive list of dependancies, as it is assumed some standard Unix tools are available on your system. Please review build error messages for more information on missing tools.
Start by cloning the repository:
git clone https://github.com/prometheus/promlens.git
cd promlens
Build PromLens:
make build
This builds the React-based web UI and compiles it into a Go binary that serves it.
Now you can run PromLens like this:
./promlens
The following documents how to run PromLens with various options and features.
Let's start with a minimal and a more elaborate example first.
To run PromLens without any advanced features (link sharing, Grafana integration, etc.) enabled:
./promlens
This starts PromLens on http://localhost:8080/ (you can override this using the --web.listen-address
flag).
To run PromLens with link sharing and Grafana integration enabled, using SQLite for link sharing (note that you will need to provide the $GRAFANA_URL
and $GRAFANA_API_KEY
environment variables for this to work):
./promlens \
--shared-links.sql.driver=sqlite \
--shared-links.sql.dsn=/tmp/promlens-links.db \
--grafana.url=$GRAFANA_URL \
--grafana.api-token=$GRAFANA_API_TOKEN
To see all available command-line flags, run:
./promlens --help
For boolean flags that default to true
, you can set them to false
by specifying --no-<flag-name>
, e.g. --no-shared-links.sql.create-tables
.
Link sharing allows persisting the state of an entire PromLens query page and sharing it with others.
By default, the link sharing backend is disabled. You can enable link sharing either via Google Cloud Storage, MySQL, or SQLite:
To use Google Cloud Storage (GCS) for link sharing, set the --shared-links.gcs.bucket=<bucket name>
flag and set the GOOGLE_APPLICATION_CREDENTIALS
environment variable to point to a JSON file containing your service account credentials (needs to have permission to create, delete, and view objects in the provided bucket).
To save shared links to a local SQLite database, set the --shared-links.sql.driver=sqlite
and --shared-links.sql.dsn=<database filename>
flags.
To save shared links in a MySQL database, set the --shared-links.sql.driver=mysql
and --shared-links.sql.dsn=<data source name>
flag (see https://github.com/go-sql-driver/mysql#dsn-data-source-name for MySQL DSN specifications).
By default, PromLens will try to auto-create the necessary tables in your MySQL database. This requires the PromLens database user to have CREATE
permissions. To turn off automatic table creation for MySQL, set the --no-shared-links.sql.create-tables
flag. If you want to create tables manually, run the following against your PromLens MySQL database:
CREATE TABLE IF NOT EXISTS link (
id INT AUTO_INCREMENT PRIMARY KEY,
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
short_name VARCHAR(11) UNIQUE,
page_state TEXT
);
CREATE TABLE IF NOT EXISTS view(
id INT AUTO_INCREMENT PRIMARY KEY,
link_id INTEGER,
viewed_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
FOREIGN KEY(link_id) REFERENCES link(id)
);
To save shared links in a Postgres database, set the --shared-links.sql.driver=postgres
and --shared-links.sql.dsn=<data source name>
flag (see https://pkg.go.dev/github.com/lib/pq#hdr-Connection_String_Parameters for Postgres DSN specifications).
By default, PromLens will try to auto-create the necessary tables in your Postgres database. This requires the PromLens database user to have CREATE
permissions. To turn off automatic table creation for Postgres, set the --no-shared-links.sql.create-tables
flag. If you want to create tables manually, run the following against your PromLens Postgres database:
CREATE TABLE IF NOT EXISTS link (
id SERIAL PRIMARY KEY,
created_at timestamptz DEFAULT now(),
short_name VARCHAR(11) UNIQUE,
page_state TEXT
);
CREATE TABLE IF NOT EXISTS view(
id SERIAL PRIMARY KEY,
link_id INTEGER,
viewed_at timestamptz DEFAULT now(),
FOREIGN KEY(link_id) REFERENCES link(id)
);
To enable selection of datasources from an existing Grafana installation, set the --grafana.url
flag to the URL of your Grafana installation, as well as either the --grafana.api-token
flag (providing an API token directly as a flag) or the --grafana.api-token-file
flag (providing an API token from a file).
To create an API token suitable for looking up datasources in Grafana:
- Head to "Configuration > API Keys" in Grafana.
- Click "Add API key".
- Give the key a descriptive name, set the "Role" to "Admin", and set its life time to the desired duration (long is recommended, as you will need to regenerate it frequently otherwise).
- Press "Add" and note down the displayed API key.
You can pre-select a specific Grafana datasource when the page loads by appending a ds
query parameter to the PromLens URL. For example, https://promlens.com/?ds=1. This works along with the q
query parameter and shared links. You can find the IDs of the datasources by visiting <PromLenURL>/api/page_config
.
You can link to a specific query without creating a persisted shared link by appending a q
query parameter to the PromLens URL. For example, https://promlens.com/?q=up directly displays and executes the query up
. For more complex shared pages, we still recommend creating a full shared page link, as this allows more control over the tree view state, as well as the selected visualization methods.
Depending on whether you use advanced features, the PromLens backend has fewer or more responsibilities:
When running the PromLens without any link sharing or Grafana integration, the backend is only used to serve static assets and parse PromQL queries into a tree view:
If you enable link sharing and/or Grafana datasource integration, the backend also stores and retrieves shared links and connects to Grafana to list datasources and proxy queries: