Resource Planning for modules

Hi Team,

We installed weaviate cluster on Openshift k8s cluster.

When we try to enable the text2vec-* modules, weviate helm creates separate pods for these modules to be enabled.

How to perform resource planning for these modules pods ? How to set cpu/memory requirements when compared to weaviate cluster pods ?


hi !

Some of those modules (eg transformers, bind, etc) will automatically set up a pod for you in k8s if using our official helm chart.

The usage of those are not covered in our doc for resource planning:

That’s because it will depend entirely on the each usage case. For example, if you have a high update/vectorize rate, etc.

So I believe it’s a case of experiment, monitor and adjust :slight_smile:

Let me know if that helps.


Thank you for the details !

1 Like