On Fri, 28 Mar 2008 12:56:12 -0500, Brian Peterson <[EMAIL PROTECTED]> wrote:
>On Fri, 28 Mar 2008 12:03:54 -0500, Mark Zelden wrote: > >>Yes. That eliminates the COMPRESS issue or the problem you might >>run into if the library takes an additional extent. The second problem >>was fixed about 15 years ago in MVS/ESA V4. JES2 recognizes this >>condition via I/O error and then will close and reopen the DD for >>all converter tasks. Actually, knowing this works makes me think >>that the "trick" mentioned an an earlier post (running jobs that point >>to a nonexistent or another PROCxx DD) could work to fix the problem >>without having to bounce JES2. But I don't know if there is some >>extra code in the fix from 15 years ago that handles that situation >>differently. >> >>Mark >>-- > >I was going to contribute to this before, to correct this statement you made, >Mark. Having had the actual hands-on experience of a JES2 proclib going >away "in anger", I know that the trick of running a job pointing to a non- >existing JES2 PROCxx DD actually works - IF THE PROCLIB DATA SETS ARE ON >THE SAME VOLUME AS BEFORE. JES2 will CLOSE and reOPEN the PROCnn >concatenation, and so long as the proclib data sets are on the exact same >VOLUME, new extent information is built and JES2 is happy. > Makes sense. I was going to write this: "The problem is, just like with the COMPRESS issue, you have to flood the system with enough jobs that every converter task gets used at the same time so it is closed and re-opened in all of them. In my case, that would probably be difficult given the speed of todays machines and the fact that I have 10 of them." The "flooding" part is always how I thought it worked (and maybe it did but has changed). See my last post: http://bama.ua.edu/cgi-bin/wa?A2=ind0803&L=ibm-main&D=1&O=D&T=0&P=214254 Maybe it would be fixed after 10 jobs where submitted in a row with the non-existing JES2 PROCxx DD. Anyway... any time this has ever happened in a shop I was at (which has probably been no more than 5 times in the past 15-20 years) I have always just bounced JES2 (after making sure the PROCLIB existed and / or was moved back to its proper location). It's quick enough and always fixes the problem. BTW, I like your requirement. 5 times was 5 times too many. No reason the ENQ can't be there and removed with an operator command when needed. Mark -- Mark Zelden Sr. Software and Systems Architect - z/OS Team Lead Zurich North America / Farmers Insurance Group - ZFUS G-ITO mailto:[EMAIL PROTECTED] z/OS Systems Programming expert at http://expertanswercenter.techtarget.com/ Mark's MVS Utilities: http://home.flash.net/~mzelden/mvsutil.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
