I should have been more thoughtful and said: Sometimes problematic, sometimes adding an extra layer of complexity, sometimes simplifying the situation.
Problematic as John pointed out batch jobs that need to regularly update the same file and constantly running into each other and restart problems. Extra complexity as for example a Rexx online-ish program that has to keep trying to ENQ the dataset and implement X number of tries or Y amount of time to keep trying before it gives up. Simplification as in batch job A that has it old and batch job B that is run unscheduled and either goes down or waits for resources. (Not sure what system parameter governs that, though.) Anyway, thanks Walt, as it is obvious there are aspects of ENQ and data integrity that I need to know better. I'm glad I asked. (-: Lindy -----Original Message----- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Walt Farrell Sent: 20. elokuuta 2007 18:44 To: [email protected] Subject: Re: ways to update PS files? On 8/20/2007 9:50 AM, Lindy Mayfield wrote: > Thanks all for the clarification. This is a synopsis up what I learned, > which was sort of what I thought: > ...snipped... > - Still, though, when you get down to it, it is still QSAM and > multi-user access (without some sort of higher level manager) would > solely depend on ENQ, which, if I'm not wrong, would be pretty lame. > I'm not sure what you mean by "lame". Properly used, ENQ will provide all that you need for data integrity. In order for two users (jobs) to have the data set allocated concurrently, they would have to specify DISP=SHR. Anyone doing an update -should- instead specify DISP=OLD, which would prevent any problems. If a single updater does specify DISP=SHR (perhaps to allow simultaneous reading) then it's not clear whether any readers would see the old or the new data. On the other hand one usually can not predict whether the updater will update before or after someone reads the record, so I don't think that matters. If two updaters try to run, specifying DISP=SHR, I believe that the second one to try the OPEN will get some flavor of S213 ABEND on any supported z/OS release. -- Walt Farrell, CISSP IBM STSM, z/OS Security Design ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
