Currently Im running a 2 node cluster in a bare metal setup. The cluster is up and running for about 8 months. This setup is installed via MCS.
3tb disk and 600gb OS
Here are some screen shots.
As you can see here 2 nodes are visible within the default rack but on the cluster utilization total cores is only 24 (should be 48 cores) and memory will be 125.6 (should be 250~256 GB) while the disk space is 6tb which seems correct.
Here are the other screenshots
NODE1 machine performance. As you can see CPU resulted to 24cores also memory is 73% which seems pretty normal
While the node2 is this
NODE2 machine performance. As you can see CPU resulted to 0 while the memory used is also 0%.
Also we noticed that other components are also not optimized, for example the spark executors. Here you can see there's only 1 single active executors (should be two).
But in Drill its perfectly fine
Hope I can get attention to this as this is a critical demo cluster.