Local Deployment¶
To try out the latest release of Browsertrix on your local machine, you'll first need to have a working Kubernetes cluster.
Installing Kubernetes¶
Before running Browsertrix, you'll need to set up a running Kubernetes cluster.
Today, there are numerous ways to deploy Kubernetes fairly easily, and we recommend trying one of the single-node options, which include Docker Desktop, microk8s, minikube, and k3s.
The instructions below assume a local package manager for your platform (eg. brew
for macOS, choco
for Windows, etc...) is already installed.
Cloning the repository at https://github.com/webrecorder/browsertrix is only needed to access additional configuration files.
Here are some environment specific instructions for setting up a local cluster from different Kubernetes vendors:
Docker Desktop (recommended for macOS and Windows)
For macOS and Windows, we recommend testing out Browsertrix using Kubernetes support in Docker Desktop as that will be one of the simplest options.
-
Install Docker Desktop if not already installed.
-
Under Settings > Kubernetes, ensure Enable Kubernetes is checked.
-
Restart Docker Desktop if asked, and wait for it to fully restart.
-
Install Helm, with
brew install helm
(macOS) orchoco install kubernetes-helm
(Windows) or following some of the other install options
MicroK8S (recommended for Ubuntu)
For Ubuntu and other Linux distros, we recommend using MicroK8S for both local deployment and production.
-
Install MicroK8s, by running
sudo snap install microk8s --classic
see more detailed instructions here or alternate installation instructions here. -
Install the following addons:
microk8s enable dns hostpath-storage registry helm3
. (For production, also addingress cert-manager
to the list of addons) -
Wait for add-ons to finish installing with
microk8s status --wait-ready
Note: microk8s comes with its own version helm, so you don't need to install it separately. Replace helm
with microk8s helm3
in the subsequent instructions below.
Minikube (Windows, macOS, or Linux)
-
Install Minikube following installation instructions, eg.
brew install minikube
. Note that Minikube also requires Docker or another container management system to be installed as well. -
Install Helm, with
!sh brew install helm
(macOS) orchoco install kubernetes-helm
(Windows) or following some of the other install options
K3S (recommended for non-Ubuntu Linux)
-
Install K3s as per the instructions
-
Install Helm, with
brew install helm
(macOS) orchoco install kubernetes-helm
(Windows) or following some of the other install options -
Set
KUBECONFIG
to point to the config for K3S:export KUBECONFIG=/etc/rancher/k3s/k3s.yaml
to ensure Helm will use the correct version.
Launching Browsertrix with Helm Repository¶
Once you have a running Kubernetes cluster with Helm 3 installed using one of the options, you can add the Browsertrix Helm Repository:
and then install the latest version of the Browsertrix helm chart with:
You can optionally specify a specific version with the --version
flag:
The versions correspond to the available Release Tags
Installing from GitHub Release Directly¶
Alternatively, you can also use Helm to install a specific version of Browsertrix directly from the latest GitHub release, if you don't wish to add the Helm repository
helm upgrade --install btrix \
https://github.com/webrecorder/browsertrix/releases/download/VERSION/browsertrix-VERSION.tgz
However, the Helm repository option is recommended as it makes upgrading to the latest version easier.
MicroK8S
If using microk8s, the commands will be:
microk8s helm3 repo add browsertrix https://docs.browsertrix.com/helm-repo/
microk8s helm3 upgrade --install btrix browsertrix/browsertrix
for the Helm repo option or, for a direct install:
microk8s helm3 upgrade --install btrix \
https://github.com/webrecorder/browsertrix/releases/download/VERSION/browsertrix-VERSION.tgz
Note: Subsequent commands will also use microk8s helm3
instead of helm
.
The default setup includes the full Browsertrix system, with frontend, backend api, db (via MongoDB), and storage (via Minio)
An admin user with name admin@example.com
and password PASSW0RD!
will be automatically created.
With Helm, additional YAML files can be added to further override previous settings.
Some possible settings can be changed are found in chart/examples/local-config.yaml.
For example, to change the default superadmin, uncomment the superadmin
block in local-config.yaml
, and then change the username (admin@example.com
) and password (PASSW0RD!
) to different values. (The admin username and password will be updated with each deployment).
To change the local port, change local_service_port
setting.
You can then redeploy with these additional settings by running:
helm upgrade --install btrix https://github.com/webrecorder/browsertrix/releases/download/VERSION/browsertrix-VERSION.tgz \
-f ./chart/examples/local-config.yaml
The above examples assumes running from a cloned Browsertrix repo, however the config file can be saved anywhere and specified with -f <extra-config.yaml>
.
Waiting for Cluster to Start¶
After running the helm command, you should see something like:
Release "btrix" does not exist. Installing it now.
NAME: btrix
LAST DEPLOYED: <time>
NAMESPACE: default
STATUS: deployed
REVISION: 1
TEST SUITE: None
After that, especially on first run, it may take a few minutes for the Browsertrix cluster to start, as all images need to be downloaded locally.
You can try running the following command to wait for all pods to be initialized:
The command will exit when all pods have been loaded, or if there is an error and it times out.
If the command succeeds, you should be able to access Browsertrix by loading: http://localhost:30870/ in your browser.
Minikube (on macOS)
When using Minikube on a macOS, the port will not be 30870. Instead, Minikube opens a tunnel to a random port,
obtained by running minikube service browsertrix-cloud-frontend --url
in a separate terminal.
Use the provided URL (in the format http://127.0.0.1:<TUNNEL_PORT>
) instead.
Debugging Pod Issues¶
If this command fails, you can also run kubectl get pods
to see the status of each of the pods.
There should be 4 pods listed: backend, frontend, minio, and mongodb. If any one is not ready for a while, something may be wrong.
To get more details about why a pod has not started, run kubectl describe <podname>
and see the latest status at the bottom.
Often, the error may be obvious, such as failed to pull an image.
If the pod is running, or previously ran, you can also get the logs from the container by running kubectl logs <podname>
The outputs of these commands are helpful when reporting an issue on GitHub
Updating the Cluster¶
To update the cluster, for example to update to new version NEWVERSION
, re-run the same command again, which will pull the latest images. In this way, you can upgrade to the latest release of Browsertrix. The upgrade will preserve the database and current archives.
helm upgrade --install btrix https://github.com/webrecorder/browsertrix/releases/download/NEWVERSION/browsertrix-NEWVERSION.tgz
Uninstalling¶
To uninstall, run helm uninstall btrix
.
By default, the database + storage volumes are not automatically deleted, run helm upgrade ...
again to restart the cluster in its current state.
If you are upgrading from a previous version, and run into issues with helm upgrade ...
, we recommend uninstalling and then re-running upgrade.
Deleting all Data¶
To fully delete all persistent data (db + archives) created in the cluster, run kubectl delete pvc --all
after uninstalling.
Deploying for Local Development¶
These instructions are intended for deploying the cluster from the latest releases published on GitHub. See setting up cluster for local development for additional customizations related to developing Browsertrix and deploying from local images.