Not isolated poor performance…
Looking out across the rest of the dasd environment quickly shows that many devices are performing poorly.
The large IOSQ time for ZPDHL8 shows that either datasets must be moved off this volume, or the data must be partitioned to spread the I/O workload across more logical volumes.
The large Disc (disconnect) times are caused by a high cache miss rate, coupled with the normally poor physical read performance from RAID devices.
The Pend times show that the channel path to the dasd is overloaded.
As a summarization, this entire dasd environment needs to be re-configured.
