Connecting a local service to a K8s cluster (Local Mode)
- Target Kubernetes workload or target k8s workload - a Kubernetes Workload deployed in a k8s cluster on the basis of the Garden service config.
- Local service - a locally deployed and running application which is supposed to replace a target Garden service configured in local mode.
This feature is still experimental. We're still working on some open tasks to improve the feature stability and usability. It means that:
- some incompatible changes can be made until the first non-experimental release
- there are some functional limitations, see the Current limitations section below
By configuring a Garden service in local mode, one can replace a target Kubernetes workload in a k8s cluster with a local service (i.e. an application running on your local machine).
Local mode feature is only supported by certain module types and providers.
Local mode uses
kubectl
port-forwarding and plain SSH port forwarding under the hood.Requirements for the local machine environment:
- OpenSSH 7.6 or higher
- Kubectl
There is a number of functional limitations in the current version.
The best matching use-case for local mode is to locally run an "isolated" service, i.e. a service that does not make any calls to other services.
If your service makes HTTP calls to some other services using k8s DNS names, then such calls will fail because the local DNS configuration is not aware about any DNS names configured in the k8s cluster.
A service cannot be running in local and dev modes simultaneously. Local mode always takes precedence over dev mode if both are configured in the relevant
garden.yml
configuration file and if both --local
and --dev
flags are enabled.The local mode is not supported natively for Windows OS. It should be used with WSL in Windows environments.
Only one container can be run in local mode for each
kubernetes
or helm
service. This limitation is planned to be removed in Garden Core 0.13
.The local mode leaves the proxy container deployed in the target k8s cluster after exit. The affected services must be re-deployed manually by using
garden deploy
.Usually, a Garden service declares a configuration and a deployment policy of a k8s service. A typical deployment flow looks like this: the Garden service takes its
Dockerfile
, builds an image if necessary, configures a Docker container, configures k8s entities and deploys them to the k8s cluster.The local mode changes the usual deployment flow. It does the following on-the-fly modifications to the target k8s cluster in the deployment phase:
- 1.The target k8s workload's container is replaced by a special proxy container which is based on openssh-server. This container exposes its SSH port and the same HTTP ports as the target Garden service.
- 2.The number of replicas of the target k8s workload is always set to
1
. - 3.The local service is started by Garden if
localMode.command
configuration option is specified in the service'sgarden.yml
. Otherwise, the local service should be started manually. - 4.The SSH port forwarding from a randomly assigned local port to the proxy container SSH port is initialized by means of
kubectl port-forward
command. - 5.The reverse port forwarding (on top of the previous SSH port forwarding) between the remote proxy container's HTTP port and the local application HTTP port is established by means of
ssh
command.
This connection schema allows to route the target k8s workload's traffic to the local service and back over the proxy container deployed in the k8s cluster. The actual service is running on a local machine, and the workload k8s service is replaced by the proxy container which connects the local service with the k8s cluster via port-forwarding.
In order to maintain secure connections, Garden generates a new SSH key pair for each service running in local mode on every CLI execution.
Note! Garden automates the SSH key acceptance with option
-oStrictHostKeyChecking=accept-new
, this is the reason why you need OpenSSH 7.6 or higher.To configure a service for local mode, add
localMode
to your module/service configuration to specify your target services.The startup, readiness and liveness probes are disabled for all services running in local mode. This has been done because of some technical reasons.
The lifecycle of a local service can be completely controlled by a user. Thus, the health checks may be unwanted and obstructing.
The k8s cluster readiness checks are applied to a proxy container which sends the traffic to the local service. When a readiness probe happens, the target local service and the relevant port forward are not ready yet. Thus, the readiness probe can cause the failure of the local mode startup.
The liveness checks can cause unnecessary re-deployment of the proxy container in the target cluster. Also, those checks create some extra traffic to the local service. That might be noisy and unnecessary if the local service is running in the debugger.
kind: Module
name: node-service
type: container
services:
- name: node-service
args: [ npm, start ]
localMode:
ports:
- local: 8090 # The port of the local app, will be used for port-forward setup.
remote: 8080 # The port of the remote app, will be used for port-forward setup.
# Starts the local app which will replace the target one in the k8s cluster.
# Optional. If not specified, then the local app should be started manually.
command: [ npm, run, serve ]
# Defines how to restart the local app on failure/exit.
# Optional. If not specified, then the default values will be applied.
restart:
delayMsec: 2000 # 2 sec delay between local app restarts
max: 100 # limit restart attempts to 100
...
kind: Module
name: backend
type: kubernetes # this example looks the same for helm modules (i.e. with `type: helm`)
localMode:
ports:
- local: 8090
remote: 8080
command: [ "../backend-local/main" ]
containerName: backend
serviceResource:
kind: Deployment
name: backend-deployment
...
# manifests or files
A
kubernetes
module example can be found in the local-mode-k8s project
. A helm
module example can be found in the local-mode-helm project
.To deploy your services with local mode enabled, you can use
deploy
or dev
commands:# Deploy specific services in local mode:
garden deploy --local=myservice
garden deploy --local=myservice,my-other-service
# Deploy all applicable services in local mode:
garden deploy --local
garden deploy --local=*
# The dev command can deploy specific services in local mode:
garden dev --local=myservice
garden dev --local=myservice,my-other-service
# The dev command can deploy all applicable services in local mode:
garden dev --local
Local mode always runs in persistent mode, it means that the Garden process won't exit until it's terminated explicitly. All port-forwards established by local mode will be stopped on the process exit. The local application will be stopped if it was started via the
localMode.command
configuration option. Otherwise, if the local application was started manually, it will continue running.If you run your local application with the
localMode.command
configuration option, then you can easily watch the local application's logs in real-time by running a garden
command with verbose
log level:garden deploy --local -l 3
# or
garden dev --local -l 3
Otherwise, you can find the logs in
.garden/deploy.debug.*.log
files.