A Kubernetes controller for configuring and managing multiple InfluxDB (+ Cloud) Instances. A control plane for managing multiple-instances of Influx; located inside and out of Kubernetes.
Currently, this is being developed locally against kind.
The current steps involved to create a cluster and install paradox are as follows:
- Create a kind cluster
kind create cluster --name paradox
. - Ensure your current context is pointed at the new
kind-paradox
context. - Run
make install
to configure the CRDs against the cluster. - Run
make run
to start the controller.
See the samples directory for some example configuration.
These will need to be adjusted to point to and authenticate against either a cloud or local Influx instance.
- Create, manage and replicate1 Influx resources via a declarative API.
- Leverage common and established k8s tooling (e.g. kubectl) and standards.
- Tightly integrate Influx resources into Kubernetes.
- Influx resource discovery via k8s APIs.
- Automate Secret management (Influx token <-> k8s Secret).
Footnotes
-
Replicate organizational and resource structure between instances (not the time-series data it-self). ↩