- Yair Mark
When you install a Helm chart or look at a chart's status the status of the different components of the chart are displayed together with a README that details how to use the chart. This README is very important as it describes to the users of your chart how exactly your chart is to be configured and used. In the absence of this you need to look at a chart's value file and then its template files to understand how exactly you need to configure it.
Today one of the charts we were trying to install had basically no README. The chart itself is awesome but we struggled to work out what exactly we had to do to get this chart to work. The chart installed and ran successfully. All components in the chart were running in Kubernetes without errors but when trying to interact with the different components we kept running into errors.
As a mob we time boxed this and found after hitting the time limit that the chart was not going to work for us. Luckily this chart also had an associated Docker container. What we ended up doing is creating the Kubernetes manifest files ourselves using this Docker container. Using this approach we had full control and knowledge of what was happening with this component in our cluster.
Helm charts can really make life easier but if the chart needs tonnes of config and does not have clear documentation on how to use the chart it can end up being faster forgoing the chart. Instead with a complex chart one approach that can be faster is to create Kubernetes manifest files yourself.