Mainframe Stuff

1.  To shut down JES3 on the local, you can issue the JES3 "*SEND" command to    route a JES3 "* RETURN " command to system B:      *t,tk4b,* return    JES3 will end on system B.  You will need to enter "QUIESCE" directly on the    system B MVS console.                   To shut down JES3 on the global processore, initiate first your standard MVS shutdown procedure. In TK4, that is                 from the JES3 console (you can enter MVS commands from the JES3 console, btw)                f bsppilot,shutdown                 Once everyhting is shutdown, you stop JES3 by issuing this command on the JES3 console:                *return               The J...

UPDATE: January 19, 2019 The procedures here work for me and hundreds of others repeatedly and I have made a video about all this which you can find on Youtube here : Prerequisites 1. Hercules 3.12/Spinhawk/4.00(hyperion) with Peter Jansen's CTC extensions I use the Hercules as provied by the latest version of Juergen Winkelmann's amazing MVS 3.8J TK4- distribution (update 8) which you can get from here.  Notice that we don't actually use the MVS3.8J distrubution at all, we only use the Hercules binaries that come with TK4-. You find them under ../hercules/linux/64/bin 2. The MVS 3.8J with JES3 distribution from Kevin's website. Notice that I have a MVS3.8A and a MVS3.8B directory. A is for the global MVS image and B is for the local MVS image. You need both TK4A:   http://www.j76.org/misc/tk4-_ update_07.jes3_2016-09-15.zip TK4B:   http://www.j76.org/misc /tk4-_update_07.jes3b_2016-10- 04.zip Or you can also get them from here :  3. an...

JES2 processing is considered to be independently controlled. Each JES2 image processes its own job input, job scheduling and job output. JES2 uses a contention-based paradigm for managing a multi-image workload. Each image “wakes up” periodically and attempts to serialize the JES2 checkpoint using a hardware serialization technique. The image that succeeds carries out job and output scheduling activities. The others go back “to sleep” waiting for another chance. There is no cooperative decision-making process. The order in which activities occur and on which images is random.   By contrast, JES3 processing is considered to be centrally controlled. One image is designated as the focal point for the entry and distribution of jobs and for the control of resources needed by the jobs. That image, called the global processor, distributes work to itself and the other images in the configuration, known as local processors. It is from the global processor that JES3 manages jobs and ...

With Peter Jansen's amazing CTCE adaptor addition to Hercules , we now have for the first time ever the ability to create JES2 MAS (multi-access spool) clusters, and,  JES3 single system image (SSI) complexes. Yes, yes, I know. Some of the "wizards" out there will say they have been doing this for years with VM and the coupling facility code in there. But I am talking about using pure Hercules and since CTCE support is still quite new, it's really not that long that we can do JES3 clusters on Hercules Thanks to Kevin Leonard 's untiring work to resurrect JES3 version 3.00 (formerly known as ASP), we now have a running MVS 3.8j with JES3 complex up and running. Here is a screenshot of a full MVS 3.8 JES3 complex up and running: Notice, how TK4A and TK4B are both online and communciating to each other via the CTC devices. That's a Global Processor and a Local Processor in a JES3 complex.  Over the next few blogs I will outline how to obtain, st...