Has someone found issues with Dremio UI latency?
In our case, Dremio is being run on an EC2 cluster well configured to guarantee High Availability, with 2 coordinators (m5.2xlarge), one in production and the other standby, and 3 executors (r5d.2xlarge). The metadata store is located on a shared volume using EFS. On top of this cluster, we set up an application load balancer with 4 minutes timeout. Sometimes, Dremio UI takes a while to respond, resulting in a 504 error.