0

I have a Kubernetes pod communicating to a neptune cluster over a transit gateway. Querying and storing data over the primary cluster endpoint works without any issues at all, however, swapping to a reader endpoint just times out. This has been happening for over 24 hours now.

What could be the root cause of this? This is a java-based application running gremlin 3.4.10 on the latest Neptune engine.

I'm getting same results locally with credentials and on remote server. Again, the primary endpoint works in both environments, but the read-only does not work at all.

Ryan
  • 1,102
  • 1
  • 15
  • 30

1 Answers1

0

The replicas were being deployed in subnets that did and did not have external access (ie, half the subnets were DMZ subnets). I updated the subnet group for the cluster/replicas.

Ryan
  • 1,102
  • 1
  • 15
  • 30