400GB as heap space for Namenode is bit high. The GC pause time will be
very high.
For a cluster with about 6PB, approx 20GB is decent memory.
As you mentioned it is HA, so it is safe to assume that the fsimage is
check pointed at regular intervals and we do not need to worry during a
manual
I am currently supporting Single Name Service is HA, Based on QJM with 0.9
PT data with 55-58 Million Object [ files + Blocks ] with 36G of JVM heap
with G1GC.
I would recommend starting with 16G and scale depending on your blocks with
G1GC Garbage collection.
Thanks,
On Fri, Aug 18, 2017 at 4