These Helm charts are design to facilitate the deployment of GXF (and depending applications) to Kubernetes.
Helm must be installed to use the charts. Please refer to Helm's documentation to get started.
Once Helm has been set up correctly, add the repo as follows:
helm repo add gxf-charts https://osgp.github.io/charts/
If you had already added this repo earlier, run helm repo update
to retrieve the latest versions of the packages. You can then run helm search repo <alias>
to see the charts.
To install the chart:
helm install my-<chart-name> <alias>/<chart-name>
To uninstall the chart:
helm delete my-<chart-name>
Some useful probes are available in GXF deployments.
You can set gxf.enableDefaultProbes
to true
to enable simple probes on the three Kubernetes probes (startup, readiness and liveness).
They will check for a text file /probe/probe.txt
on Tomcat, which will only be served when Tomcat is started.
When you use Prometheus metrics the GXF way (specifying the port through property metrics.prometheus.port
),
use that instead of the default probe, because it actually checks if your app is running. An example config for this situation:
gxf:
livenessProbe:
failureThreshold: 6
periodSeconds: 5
httpGet:
port: 9404
path: /metrics
When you run multiple WARs in your container, you can also use the supplied probe script /usr/local/bin/probe.sh
which tries to contact all Prometheus endpoints to see if the individual WARs are available. An example configuration for this situation:
gxf:
livenessProbe:
failureThreshold: 3
periodSeconds: 10
exec:
command:
- /bin/bash
- '-c'
- source /usr/local/bin/probe.sh