This monitor tracks the aggregated disk throughput of the whole cluster. The value is calculated by adding the disk throughput values, in KBps, of all nodes.
Sustained high disk throughput is usually caused by jobs that perform excessive disk paging. These jobs usually slow down because reading from and writing to disks is slower than reading from and writing to memory.
Identify the compute nodes with higher disk throughput while executing jobs and check their memory usage. If the memory usage is high, consider adding more memory to that compute node.