@kaniz - After changing my resource group/cluster I was able to execute the query successfully.

As you said the issue was due to limited resources available in the cluster.