________________________________ From: Tom Wasik <[email protected]> To: [email protected] Sent: Fri, June 18, 2010 12:23:05 PM Subject: Re: JES2 Queue Times / SMF30 Data Zones
Let me explain what each time means/comes from and perhaps that will help answer your questions: SMF30RST - The comment in the IFASMFR3 macro states "TIME, IN HUNDREDTHS OF A SECOND READER RECOGNIZED THE JOB CARD FOR THIS JOB." This is precisely correct. It is when JES2 input processing sees the JOB card. In z/OS 1.7 this happens a bit earlier for internal readers than prior releases (in case that matters). 11.08.15 JOB94550 ---- TUESDAY, 01 JUN 2010 ---- This is when the first message is placed in the job log for the job. This is when the data set is created. In general, the first message for a job is the RACF message that the job has been authenticated and assigned a userid. This occurs at the end of input processing (when JES2 realizes that it has a complete job). For TSO SUBMIT processing, it is when it reaches the end of the job stream and issues an ENDREQ macro to get the job id. It also occurs when a IEBGENER closes the internal reader. Or it occurs when the next JOB card is encountered. The difference between these 2 time stamps is the time it took to complete input processing for the job. This includes the RACF check and writing the job to SPOOL. It also includes the time the submitting application took to actually write the job. There have been cases when an application was submitting multiple jobs but did not tell JES when one job ended. In that case, the job will sit there until the next job is written to the internal reader. There are some rare cases where the first message into a job log does not occur during input processing. For example when you supply a userid and password on the job card and did not enable the "ICH70002I YOUR PASSWORD WILL EXPIRE IN xx DAYS." message. Then the first message occurs at conversion time. Just submit a job with TYPRUN=JCLHOLD to see what messages are issued at input time vs conversion time. ---------------SNIP---------- Good information Tom, thanks. I was running into similar issues a long time ago. But in slightly different circumstances. What gets interesting in an NJE environment. The timing is unusual and while semi explainable (for the most part) what was unexplainable (at least to me) when date lines get crossed and it seemed at that time I was coming up with negative times. I never did get a good explanation as to why although I think I read an IBM article about NJE and they did address it (negative times) anyway its been ages and it might not be an issue now as it was back then. Ed ---------------------------------------------------------------------- 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
