Are you about to run out of ECSA or CSA on any of your LPARs?
deanab 1200006JV8 Visits (12064)
Why monitor CSA and ECSA? Its simple, when common storage (CSA, ECSA, SQA, and ESQA) is exhausted, a system outage will occur. When ESQA is exhausted, further requests for ESQA are allocated from ECSA. When ECSA is exhausted, further requests are allocated from the much smaller CSA storage. The main reasons why common storage shortages occur:
Noone can afford to watch a screen all day, issuing display commands and looking for when the system is about to run out of CSA storage. A best practice is to create a situation that detects when a shortage occurs, before the shortage causes a system outage.
OMEGAMON XE on z/OS provides two sample situations, one that emits a warning situation alert when a system has reached 80% allocation and a second that emits a critical situation alert when a system has exceeded 90% allocation. There are differing philosophies on whether running one or two situations is best. Although running two situations has the advantage of providing a 2-stage response - a warning stage, causing systems programmers to investigate, and then a critical stage, allowing automated action to avoid the system outage. Running a single situation uses less CPU and relies on your experts to find the correct resolution; often its too risky to automate cancelling address spaces, even to save a system.
1) Operations calls Jim, a z/OS systems programmer, to investigate a situation alert. The situation is reporting critical allocation percent of ECSA on LPARs SP11 and SP13. Jim views the situation alerts in the OMEGAMON Enhanced 3270 User Interface. He navigates to the ITM Situation Status workspace.
2) Jim navigates to the Enterprise Sysplex Overview workspace and notes the ECSA In User Percent exception is active on CANSP11.
3) He types the 'T' action character on that row to navigate to the Storage Overview workspace. ECSA continues to be above 90%, a critical level.
4) Jim uses default navigation from the Common Storage sub-panel to navigate to the Storage Usage by Address Space workspace.
5) Jim identifies some test address spaces holding ECSA storage and stops or cancels them. He has saved the LPAR from a system outage.