Submitting security fixes is something that may put the community at risk. Please read through our [security policy](SECURITY.md) and submit vulnerabilities via an [advisory] + [issue] instead. We encourage you to submit how to fix a vulnerability if you know how to, this is not required. Following the security policy allows us to properly test, fix bugs. This review allows us to notice, if there are any changes necessary to unrelated parts like the documentation. [**PLEASE SEE OUR SECURITY POLICY.**](SECURITY.md) [advisory]: https://github.com/louislam/uptime-kuma/security/advisories/new [issue]: https://github.com/louislam/uptime-kuma/issues/new?template=security_issue.yml
If you come across a bug and think you can solve, we appreciate your work. Please make sure that you follow these rules: - keep the PR as small as possible, fix only one thing at a time => keeping it reviewable - test that your code does what you claim it does. Because maintainer time is precious, junior maintainers may merge uncontroversial PRs in this area.
Please add **all** strings that are translatable to `src/lang/en.json`. If translation keys are omitted, they cannot be translated. **Do not include any other languages in your initial pull request** (even if it is your mother tongue) to avoid merge conflicts between Weblate and `master`. Once your PR is merged into `master`, the strings can be translated by awesome people donating their language skills.
We use Weblate to localise this project into many languages. If you want to help translate Uptime Kuma into your language, please see [these instructions on how to translate using Weblate](https://github.com/louislam/uptime-kuma/blob/master/src/lang/README.md).
There are some cases where a change cannot be done directly in Weblate and requires a PR:
- A text may not yet be localisable. In this case, **adding a new language key** via `{{ $t("Translation key") }}` or [`
To set up a new notification provider these files need to be modified/created:
- `server/notification-providers/PROVIDER_NAME.js` is where the heart of the
notification provider lives.
- Both `monitorJSON` and `heartbeatJSON` can be `null` for some events. If
both are `null`, this is a general testing message, but if just
`heartbeatJSON` is `null` this is a certificate expiry.
- Please wrap the axios call into a
```js
try {
let result = await axios.post(...);
if (result.status === ...) ...
} catch (error) {
this.throwGeneralAxiosError(error);
}
```
- `server/notification.js` is where the backend of the notification provider
needs to be registered. _If you have an idea how we can skip this step, we
would love to hear about it ^^_
- `src/components/NotificationDialog.vue` you need to decide if the provider
is a regional or a global one and add it with a name to the respective list
- `src/components/notifications/PROVIDER_NAME.vue` is where the frontend of
each provider lives. Please make sure that you have:
- used `HiddenInput` for secret credentials
- included all the necessary helptexts/placeholder/.. to make sure the
notification provider is simple to setup for new users. - include all
translations (`{{ $t("Translation key") }}`,
[`i18n-t keypath="Translation key">`](https://vue-i18n.intlify.dev/guide/advanced/component.html))
in `src/lang/en.json` to enable our translators to translate this
- `src/components/notifications/index.js` is where the frontend of the
provider needs to be registered. _If you have an idea how we can skip this
step, we would love to hear about it ^^_
Offering notifications is close to the core of what we are as an uptime
monitor. Therefore, making sure that they work is also really important.
Because testing notification providers is quite time intensive, we mostly
offload this onto the person contributing a notification provider.
To make sure you have tested the notification provider, please include
screenshots of the following events in the pull-request description:
- `UP`/`DOWN`
- Certificate Expiry via
Using the following way to format this is encouraged:
```md
| Event | Before | After |
| ------------------ | --------------------- | -------------------- |
| `UP` |  |  |
| `DOWN` |  |  |
| Certificate-expiry |  |  |
| Domain-expiry |  |  |
| Testing |  |  |
```
Because maintainer time is precious, junior maintainers may merge
uncontroversial PRs in this area.
To set up a new notification provider these files need to be modified/created:
- `server/monitor-types/MONITORING_TYPE.js` is the core of each monitor.
The `async check(...)`-function should:
- in the happy-path: set `heartbeat.msg` to a successful message and set `heartbeat.status = UP`
- in the unhappy-path: throw an `Error` for each fault that is detected with an actionable error message.
- NEVER set `heartbeat.status = DOWN` unless you want to explicitly ignore retries.
- `server/uptime-kuma-server.js` is where the monitoring backend needs to be
registered. _If you have an idea how we can skip this step, we would love to
hear about it ^^_
- `src/pages/EditMonitor.vue` is the shared frontend users interact with.
Please make sure that you have: - used `HiddenInput` for secret
credentials - included all the necessary helptexts/placeholder/.. to make
sure the notification provider is simple to setup for new users. - include
all translations (`{{ $t("Translation key") }}`,
[`
be sure to **create an empty draft pull request or open an issue, so we can have a discussion first**. This is especially important for large pull requests or when you don't know if it will be merged or not. When adding new features, please also add tests to ensure your changes work as expected and to prevent future regressions. Because of the large impact of this work, only senior maintainers may merge PRs in this area.
Contributing is easy and fun. We will guide you through the process:
1. **Fork** the [Uptime-Kuma repository](https://github.com/louislam/uptime-kuma/) and **clone** it to your local machine.
2. **Create a new branch** for your changes (e.g., `signal-notification-provider`).
3. **Make your changes** and **commit** them with a clear message.
4. **Push** your changes to your forked repository.
5. **Open a pull request** to the `master` branch of the Uptime Kuma repository.
- For large changes, please open a **draft pull request** first to discuss the changes with the maintainers.
6. **Provide a clear and concise description** of the changes you've made and link any related issues.
7. **Complete the PR checklist** and make sure all CI checks pass.
8. **Request a review** when your pull request is ready.
## When Can You Change the PR Status to "Ready for Review"?
A PR should remain in **draft status** until all tasks are completed.
Only change the status to **Ready for Review** when:
- You have implemented all planned changes.
- Your code is fully tested and ready for review.
- You have updated or created the necessary tests.
- You have verified that CI/CD checks pass successfully.
A volunteer maintainer will review your PR as soon as possible.
You can help us by reviewing other PRs or taking a look at open issues.
## The following rules are essential for making your PR mergeable
- Merging multiple issues by a huge PR is more difficult to review and causes
conflicts with other PRs. Please
- (if possible) **create one PR for one issue** or
- (if not possible) **explain which issues a PR addresses and why this PR
should not be broken apart**
- Make sure your **PR passes our continuous integration**. PRs will not be
merged unless all CI-Checks are green.
- **Breaking changes** (unless for a good reason and discussed beforehand) will
not get merged / not get merged quickly. Such changes require a major version
release.
- **Test your code** before submitting a PR. Buggy PRs will not be merged.
- Make sure the **UI/UX is close to Uptime Kuma**.
- **Think about the maintainability**: Don't add functionality that is
completely **out of scope**. Keep in mind that we need to be able to maintain
the functionality.
- Don't modify or delete existing logic without a valid reason.
- Don't convert existing code into other programming languages for no reason.
### Continuous Integration
All pull requests must pass our continuous integration checks. These checks include:
- **Linting**: We use ESLint and Stylelint for code quality checks. You can run the linter locally with `npm run lint`.
- **Formatting**: We use Prettier for code formatting. You can format your code with `npm run fmt` (or CI will do this for you)
- **Testing**: We use Playwright for end-to-end tests and have a suite of backend tests. You can run the tests locally with `npm test`.
I ([@louislam](https://github.com/louislam)) have the final say.
If your pull request does not meet my expectations, I will reject it, no matter how much time
you spent on it.
We will assign your pull request to a [milestone](https://github.com/louislam/uptime-kuma/milestones), if we plan to review and merge it.
Please don't rush or ask for an ETA.
We have to understand the pull request, make sure it has no breaking changes and stick to the vision of this project, especially for large pull requests.
## I'd Like to Work on an Issue. How Do I Do That?
We have found that assigning people to issues is unnecessary management
overhead. Instead, a short comment stating that you want to work on an issue is
appreciated, as it saves time for other developers. If you encounter any
problems during development, feel free to leave a comment describing what you
are stuck on. We are here to help.
## Project Style
I personally do not like something that requires a lot of configuration before
you can finally start the app. The goal is to make the Uptime Kuma installation
as easy as installing a mobile app.
- Easy to install for non-Docker users
- no native build dependency is needed (for `x86_64`/`armv7`/`arm64`)
- no extra configuration and
- no extra effort required to get it running
- Single container for Docker users
- no complex docker-compose file
- mapping the volume and exposing the port should be the only requirements
- Settings should be configurable in the frontend. Environment variables are
discouraged, unless it is related to startup such as `DATA_DIR`
- Easy to use
- The web UI styling should be consistent and nice
## Coding Styles
- 4 spaces indentation
- Follow `.editorconfig`
- Follow ESLint
- Methods and functions should be documented with JSDoc
## Name Conventions
- Javascript/Typescript: camelCaseType
- SQLite: snake_case (Underscore)
- CSS/SCSS: kebab-case (Dash)
## Tools
- [`Node.js`](https://nodejs.org/) >= 20.4.0
- [`npm`](https://www.npmjs.com/) >= 9.3
- [`git`](https://git-scm.com/)
- IDE that supports [`ESLint`](https://eslint.org/) and EditorConfig (I am using
[`IntelliJ IDEA`](https://www.jetbrains.com/idea/))
- A SQLite GUI tool (f.ex.
[`SQLite Expert Personal`](https://www.sqliteexpert.com/download.html) or
[`DBeaver Community`](https://dbeaver.io/download/))
## Git Branches
- `master`: 2.X.X development. If you want to add a new feature, your pull
request should base on this.
- `1.23.X`: 1.23.X development. If you want to fix a bug for v1 and v2, your
pull request should base on this.
- All other branches are unused, outdated or for dev.
## Install Dependencies for Development
```bash
npm ci
```
## Dev Server
We can start the frontend dev server and the backend dev server in one command.
Port `3000` and port `3001` will be used.
```bash
npm run dev
```
But sometimes you may want to restart the server without restarting the
frontend. In that case, you can run these commands in two terminals:
```bash
npm run start-frontend-dev
npm run start-server-dev
```
## Backend Server
It binds to `0.0.0.0:3001` by default.
The backend is an `express.js` server with `socket.io` integrated. It uses
`socket.io` to communicate with clients, and most server logic is encapsulated
in the `socket.io` handlers. `express.js` is also used to serve:
- as an entry point for redirecting to a status page or the dashboard
- the frontend built files (`index.html`, `*.js`, `*.css`, etc.)
- internal APIs of the status page
### Structure in `/server/`
- `jobs/` (Jobs that are running in another process)
- `model/` (Object model, auto-mapping to the database table name)
- `modules/` (Modified 3rd-party modules)
- `monitor_types/` (Monitor Types)
- `notification-providers/` (individual notification logic)
- `routers/` (Express Routers)
- `socket-handler/` (Socket.io Handlers)
- `server.js` (Server entry point)
- `uptime-kuma-server.js` (UptimeKumaServer class, main logic should be here,
but some still in `server.js`)
## Frontend Dev Server
It binds to `0.0.0.0:3000` by default. The frontend dev server is used for
development only.
For production, it is not used. It will be compiled to `dist` directory instead via `npm run build`.
You can use Vue.js devtools Chrome extension for debugging.
### Frontend Details
Uptime Kuma Frontend is a single page application (SPA). Most paths are handled
by Vue Router.
The router is in `src/router.js`
Most data in the frontend is stored at the root level, even though the router can navigate to different pages.
The data and socket logic are in `src/mixins/socket.js`.
## Database Migration
See:
- amd64, armv7 using local.
- arm64 using remote arm64 cpu, as the emulator is too slow and can no longer
pass the `npm ci` command.
1. Add the public key to the remote server.
2. Add the remote context. The remote machine must be arm64 and installed
Docker CE.
```bash
docker context create oracle-arm64-jp --docker "host=ssh://root@100.107.174.88"
```
3. Create a new builder.
```bash
docker buildx create --name kuma-builder --platform linux/amd64,linux/arm/v7
docker buildx use kuma-builder
docker buildx inspect --bootstrap
```
4. Append the remote context to the builder.
```bash
docker buildx create --append --name kuma-builder --platform linux/arm64 oracle-arm64-jp
```
5. Verify the builder and check if the builder is using `kuma-builder`.
`docker buildx inspect kuma-builder docker buildx ls`
1. Draft a release note
2. Make sure the repo is cleared
3. If the healthcheck is updated, remember to re-compile it:
`npm run build-docker-builder-go`
4. `npm run release-final` with env vars: `VERSION` and `GITHUB_TOKEN`
5. Wait until the `Press any key to continue`
6. `git push`
7. Publish the release note as `1.X.X`
8. Press any key to continue
9. Deploy to the demo server: `npm run deploy-demo-server`
These Items need to be checked:
- [ ] Check all tags is fine on
1. Draft a release note, check `This is a pre-release`
2. Make sure the repo is cleared
3. `npm run release-beta` with env vars: `VERSION` and `GITHUB_TOKEN`
4. Wait until the `Press any key to continue`
5. Publish the release note as `1.X.X-beta.X`
6. Press any key to continue
**Setup Repo**
```bash
git clone https://github.com/louislam/uptime-kuma-wiki.git
cd uptime-kuma-wiki
git remote add production https://github.com/louislam/uptime-kuma.wiki.git
```
**Push to Production Wiki**
```bash
git pull
git push production master
```
```bash
git rebase --onto Set up a Docker Builder (click to expand)
Release (click to expand)
Release Beta (click to expand)
Release Wiki (click to expand)
Change the base of a pull request such as
master to 1.23.X (click to expand)