Monitoring Query Performance on Weaviate Cloud

Description

Server Setup Information

  • Weaviate Server Version: Weaviate Cloud
  • Deployment Method: Weaviate Cloud
  • Multi Node? Number of Running Nodes: 1
  • Client Language and Version: python connector 4.11.2
  • Multitenancy?:

Any additional Information

We are populating our own vector databases into Weaviate Cloud and using a python connector to query against it. Is there a way to monitor the process and extract more information like querying time?

There is a Monitoring module by Prometheus, but is it possible to set it up on Weaviate Cloud?

hi @louisja1 !!

Welcome to our community :hugs:

We do not expose those informations for our serverless cluster yet.

Exposing prometheus metrics directly can be done for enterprise and bring your own cloud plans.

It is important to note that all clusters hosted in our cloud are actively auto scaled and monitored by our team 24/7.

If you ever need any help, or for example, you know you are going to ingest a lot of data in a short period of time, you can reach us at any time by opening a support ticket.

We can then provision your cluster with the necessary resource so the first import runs smoothly.

Let me know if that helps!

Thanks!

Is there an easy way to get the query time without accounting round-trip network overhead between our machine and Weaviate cloud?

hi @louisja1 !!

For that you would need to run a pod inside the cluster and do the query.

Or, replicate the dataset on a local/on-prem cluster and run the query against it.

So this is possible for serverless, but It can certainly be done for bring your own cloud plans, as you have access to your cluster.

Let me know if that helps!

Thanks!

1 Like