high utilization caused by memory movement between pools. cache memory allocator errors Cache pool 0 thread

High utilization was being caused on the Netware server due to memory constantly being mapped between the pools and never settling down. This was due to nss constantly mapping the memory for its cache. The garbage collection interval was still set at the default of 5 minutes. The problem appears to be when the machine has less physical ram than the FCMS line. The OS is therefore always trying to move memory down to the L=P pool. One of the symptoms would be that the cache buffers would rapidly disappear and then reappear as the the vm pool mapped the memory back down to the L=P pool.