The DB Operator eases the pain of managing PostgreSQL and MySQL instances for applications running in Kubernetes. The Operator creates databases and make them available in the cluster via Custom Resource. It is designed to support the on demand creation of test environments in CI/CD pipelines.
DB Operator provides following features:
- Create/Delete databases on the database server running outside/inside Kubernetes by creating
Database
custom resource; - Create Google Cloud SQL instances by creating
DbInstance
custom resource; - Automatically create backup
CronJob
with defined schedule (limited feature);
- How it works - a general overview and definitions
- Creating Instances - make database instances available for the operator
- Creating Databases - creating databases in those instances
- Enabling regular Backup - and schedule cronjob
- Upgrade guide - breaking changes and guide for the upgrade
The repository contains helm charts for db-operator is moved to https://github.com/db-operator/charts New chart after db-operator > 1.2.7, db-instances > 1.3.0 will be only available in new repository.
Now both DbInstance
and Database
resources are upgraded to v1beta1
.
In case you were using connectionStringTemplate
, make sure I've migrated to secretsTempaltes
before upgrading, because connectionStringTemplate
is removed in this version. Everything else should go seamless. We've added Webhooks
that will take care of resources with an old API version and convert them to the newer one.
Installing older version of charts is still possible. Check available versions by following command.
$ helm repo add kloeckneri-old https://kloeckner-i.github.io/db-operator/
$ helm search repo kloeckneri-old/ --versions
$ helm repo add db-operator https://db-operator.github.io/charts/
$ helm install --name my-release db-operator/db-operator
To see more options of helm values, see chart repo
To see which version is working together check out our version matrix.
- go 1.15+
- docker
- make
- kubectl v1.14+ (< v1.21)
- helm v3.0.2+
- k3d
To have kubernetes environment locally, you need to install k3d.
addexamples add examples via kubectl create -f examples/
build build db-operator docker image
controller-gen Download controller-gen locally if necessary.
generate generate supporting code for custom resource types
help show this help
k3d_image rebuild the docker images and upload into your k3d cluster
k3d_install install k3d cluster locally
k3d_setup install k3d and import image to your k3d cluster
k3s_mac_deploy build image and import image to local lima k8s
k3s_mac_image import built image to local lima k8s
k3s_mac_lima_create create local k8s using lima
k3s_mac_lima_start start local lima k8s
lint lint go code
manifests generate custom resource definitions
test run go unit test
vet go vet to find issues
rebuild CRD manifests
$ make manifests
rebuild local docker image
$ make build
$ make k3d_setup
or
$ make k3s_mac_lima_create
$ make k3d_build
or
$ make k3s_mac_image
helm repo add db-operator https://db-operator.github.io/charts
helm repo update
helm upgrade my-release db-operator/db-operator --set image.repository=my-db-operator --set image.tag=1.0.0-dev --set image.pullPolicy=IfNotPresent --install
$ make test