ClickHouse
You’ll get a single-node ClickHouse analytics database in an environment.
Choose this when
- Apps need ClickHouse for analytics / OLAP
- You prefer the catalog flow over the dedicated Databases path (see chooser if unsure)
What you get
- One ClickHouse replica in your environment (no Keeper, no cluster)
- Internal HTTP (
8123) and native (9000) endpoints - Optional public HTTP via public URL or NodePort
Options
| Option | Purpose |
|---|---|
| Environment | Where ClickHouse runs |
| Access | Prefer internal-only; public HTTP only when you need it |
| Disk / storage | Data capacity |
| CPU / memory | ClickHouse is memory-heavy — size accordingly |
Connect
Copy hosts from the database connection section.
- HTTP (in cluster):
clickhouse-{app-name}.{environment}.svc:8123 - Native (in cluster):
clickhouse-{app-name}.{environment}.svc:9000 - From outside:
{cluster-public-ip}:{node-port}on HTTP only
Ingress is HTTP only. Native protocol is not exposed through nginx.
After deploy
This release is single node. Size disk and memory before production. Also see ClickHouse databases. Full deploy steps: How to use ready-made apps.