Perses is part of the CoreDash community. It belongs to the Linux Foundation. At a later stage, we want to promote the project to the Cloud Native Computing Foundation and be part of the monitoring tools like Prometheus or Thanos.
Perses is going to tackle multiple different goals:
- It aims to become a standard dashboard visualization tool for Prometheus and other datasources. It will focus on being GitOps-compatible and thus enabling a smooth "dashboards as code" workflow via a new and well-defined dashboard definition model.
- While becoming another visualization tool, Perses also aims to provide different npm packages, so it can benefit anyone that would like to embed charts and dashboards in their UI. For example, these packages might be used to improve the display of the data in the Prometheus UI.
- It also aims to offer a Kubernetes-native mode in which dashboard definitions can be deployed into and read from individual application namespaces (Using CRDs). For more information you can take a look at the doc that would give you an idea of how it would work.
- To be friendly to dashboard as code users, we want to provide a complete static validation of the dashboard format.
That means you will be able to validate your dashboard in a CI/CD using the Perses CLI (named
percli
) - The architecture should support plugins (at least for the panels)
While we already released a certain amount of versions, Perses is still in an early alpha stage and still work in progress. The current pieces that are in place are:
- The Plugin architecture has finally reached a stable point.
- The plugins concern the Variables, Panels, Queries and Datasources definitions.
- To provide a good static validation, the backend is using multiple Cue schemas and the CLI has the
lint
command. All schemas are available in the schemas folder.
- A backend REST API provides R/W access to dashboard and datasource definitions.
- A CLI that can be used to interact with the REST API. A short docs is available here
- While the UI is still in progress, we already have:
- a beginning of navigation that will help to move from a dashboard to another.
- a support of the following panel types:
- Time series charts.
- Gauge panels.
- Stat panels (single value with sparkline).
- Markdown panels (as an alternative to the Text panel)
- Visual editing of the dashboard is available, which includes a number of editable panel options
- A migration script that will help to move from Grafana to Perses is on going.
- The dashboard data model is still evolving along with the dashboard implementation and new requirements. Before reaching a stable state regarding the data model, we are waiting for feedback to know if we need to adjust and potentially break things.
Here is a not ordered list of what it can come in the future in Perses:
- Perses native on Kubernetes using CRDs
- Traces Visualization support
- Docs, a lot of docs :)
- Generating Panel #200
- Sub folder management #183
- Datasource discovery #74
There are various ways of installing Perses.
Precompiled binaries for released versions are available in the GitHub release. Using the latest release binary is the recommended way of installing Perses.
Docker images are available on Docker Hub.
You can launch a Perses container for trying it out with:
docker run --name perses -d -p 127.0.0.1:8080:8080 persesdev/perses
To build Perses from source code, You need:
- Go version 1.18 or greater.
- NodeJS version 18 or greater.
- npm version 8 or greater.
Start by cloning the repository:
git clone https://github.com/perses/perses.git
cd perses
Then you can use make build
that would build the web assets and then Perses itself (and also the Perses CLI that can
be used to interact directly with the Perses API in case you prefer to browser the API using a terminal).
make build
./bin/perses --config=your_config.yml
General instructions about how you can contribute to Perses are available in the document CONTRIBUTING.md.
If you are primarily interested in contributing to the UI application and libraries, please refer to the UI Readme. It includes quick start instructions for how to build, run, and test the React UI. It also includes details about the architecture and guidelines for development.
The code is licensed under an Apache 2.0 license.