When we do an LDTE display, we get a list of over 34,000 datasets. Of these more than 30,000 are for datasets that start with a specific prefix that show individual dataset names tracked by a job with that same prefix. They show individual dataset names that were being tracked by a DSTRIG Wob in an application. A new entry is created under the LDTE display every time an new dataset is created. Should I be concerned about this buildup? Will it affect ESP performance and what is the impact on the INDEX dataset?
The LDTE command displays DSTRIG cells in ECSA, both active and inactive. Any job in the system can examine these DSTRIG cells at any time. Therefore ESP does not free ECSA DSTRIG cells when they cease to be active, as there could be a potential system integrity issue. There is a gradual accumulation of ECSA usage when newly defined DSTRIG cells are added. DSTRIG cells vary in length between approximately 24 bytes and 92 bytes. The primary determining factor is the DSNAME length. The average length of an ECSA DSTRIG cell should be roughly 40 bytes. Therefore the accumulation of ECSA DSTRIG cells is unlikely to cause a serious ECSA usage problem. An IPL clears the ECSA DSTRIG cell queues.