This document describes how to use Tilt with any cluster for a simplified workflow that offers easy deployments and rapid iterative builds.
This setup allows for continuing deployment of the Velero server and, if specified, any provider plugin or the node-agent daemonset. It does this work by:
docker_build to live update any binary into the container/init container and trigger a re-start
Tilt will look for configuration files under
velero/tilt-resources. Most of the
files in this directory are gitignored so you may configure your setup according to your needs.
Note: To properly configure any plugin you use, please follow the plugin’s documentation.
Copy all sample files under
velero_v1_backupstoragelocation.yaml file, and the
cloud file for the storage credentials/secret.
Create a configuration file named
tilt-settings.json and place it in your local copy of
you may copy and paste the sample file found in
Here is an example:
default_registry (String, default=”"): The image registry to use if you need to push images. See the Tilt *documentation for more details.
provider_repos (ArrayString, default=): A list of paths to all the provider plugins you want to make changes to. Each provider must have a
tilt-provider.json file describing how to build the provider.
enable_providers (ArrayString, default=): A list of the provider plugins to enable. See provider plugins for more details. Note: when not making changes to a plugin, it is not necessary to load them into Tilt: an existing image and version might be specified in the Velero deployment instead, and Tilt will load that.
allowed_contexts (Array, default=): A list of kubeconfig contexts Tilt is allowed to use. See the Tilt documentation on * allow_k8s_contexts for more details. Note: Kind is automatically allowed.
use_node_agent (Bool, default=false): Indicate whether to deploy the node-agent Daemonset. If set to
true, Tilt will look for a
containing the configuration of the Velero node-agent DaemonSet.
create_backup_locations (Bool, default=false): Indicate whether to create one or more backup storage locations. If set to
true, Tilt will look for a
containing at least one configuration for a Velero backup storage location.
setup-minio (Bool, default=false): Configure this to
true if you want to configure backup storage locations in a Minio instance running inside your cluster.
enable_debug (Bool, default=false): Configure this to
true if you want to debug the velero process using
debug_continue_on_start (Bool, default=true): Configure this to
true if you want the velero process to continue on start when in debug mode. See
Delve CLI documentation.
All needed Kubernetes resource files are provided as ready to use samples in the
velero/tilt-resources/examples directory. You only have to move them to the
Because the Velero Kubernetes deployment as well as the node-agent DaemonSet contain the configuration for any plugin to be used, files for these resources are expected to be provided by the user so you may choose which provider plugin to load as a init container. Currently, the sample files provided are configured with all the plugins supported by Velero, feel free to remove any of them as needed.
For Velero to operate fully, it also needs at least one backup storage location. A sample file is provided that needs to be modified with the specific configuration for your object storage. See the next sub-section for more details on this.
You will have to configure the
velero/tilt-resources/velero_v1_backupstoragelocation.yaml with the proper values according to your storage provider. Read the
to learn what field/value pairs are required for your particular provider’s backup storage location configuration.
Below are some ways to configure a backup storage location for Velero.
Follow the provider documentation to provision the storage. We have a list of all known object storage providers with corresponding plugins for Velero.
Note: to use MinIO as an object storage, you will need to use the
AWS plugin, and configure the storage location with the
spec.provider set to
aws and the
spec.config.region set to
Here are two ways to use MinIO as the storage:
As a MinIO instance running inside your cluster (don’t do this for production!)
tilt-settings.json file, set
"setup-minio": true. This will configure a Kubernetes deployment containing a running
instance of MinIO inside your cluster. There are
necessary to expose MinIO outside the cluster.
To access this storage, you will need to expose MinIO outside the cluster by forwarding the MinIO port to the local machine using kubectl port-forward -n svc/minio 9000. Update the BSL configuration to use that as its “public URL” by adding
publicUrl: http://localhost:9000 to the BSL config. This is necessary to do things like download a backup file.
Note: with this setup, when your cluster is terminated so is the storage and any backup/restore in it.
As a standalone MinIO instance running locally in a Docker container
See these instructions to run MinIO locally on your computer, as a standalone as opposed to running it on a Pod.
Please see our locations documentation to learn more how backup locations work.
Whatever object storage provider you use, configure the credentials for in the
velero/tilt-resources/cloud file. Read the
to learn what field/value pairs are required for your provider’s credentials. The Tilt file will invoke Kustomize to create the secret under the hard-coded key
secret.cloud-credentials.data.cloud in the Velero namespace.
There is a sample credentials file properly formatted for a MinIO storage credentials in
If you would like to debug the Velero process, you can enable debug mode by setting the field
true in your
This will enable you to debug the process using
By enabling debug mode, the Velero executable will be built in debug mode (using the flags
-gcflags="-N -l" which disables optimizations and inlining), and the process will be started in the Velero deployment using
The debug server will accept connections on port 2345 and Tilt is configured to forward this port to the local machine.
Once Tilt is
running and the Velero resource is ready, you can connect to the debug server to begin debugging.
To connect to the session, you can use the Delve CLI locally by running
dlv connect 127.0.0.1:2345. See the
Delve CLI documentation for more guidance on how to use Delve.
Delve can also be used within a number of
editors and IDEs.
By default, the Velero process will continue on start when in debug mode.
This means that the process will run until a breakpoint is set.
You can disable this by setting the field
false in your
When this setting is disabled, the Velero process will not continue to run until a
continue instruction is issued through your Delve session.
When exiting your debug session, the CLI and editor integrations will typically ask if the remote process should be stopped. It is important to leave the remote process running and just disconnect from the debugging session. By stopping the remote process, that will cause the Velero container to stop and the pod to restart. If backups are in progress, these will be left in a stale state as they are not resumed when the Velero pod restarts.
To launch your development environment, run:
This will output the address to a web browser interface where you can monitor Tilt’s status and the logs for each Tilt resource. After a brief amount of time, you should have a running development environment, and you should now be able to create backups/restores and fully operate Velero.
tilt down after exiting out of Tilt
will delete all resources specified in the Tiltfile.
Tip: Create an alias to
velero/_tuiltbuild/local/velero and you won’t have to run
make local to get a refreshed version of the Velero CLI, just use the alias.
Please see the documentation for how Velero works.
A provider must supply a
tilt-provider.json file describing how to build it. Here is an example:
Each provider plugin configured to be deployed by Velero’s Tilt setup has a
live_reload_deps list. This defines the files and/or directories that Tilt
should monitor for changes. When a dependency is modified, Tilt rebuilds the provider’s binary on your local
machine, copies the binary to the init container, and triggers a restart of the Velero container. This is significantly faster
than rebuilding the container image for each change. It also helps keep the size of each development image as small as
possible (the container images do not need the entire go toolchain, source code, module dependencies, etc.).