Having any issue on z/dt with processors. The existing processors/prcsload copied over from z/os work fine. When a re-compiled a few to pull in an updated processor include, it broke LMAP. Of course I can recompile on z/os without issue. Just trying to determine which puzzle piece is missing. I'm still comparing TRMOD traces, but wondering if anyone else has run into this so the sysprogs can adjust the "Gold Copy" copy process accordingly.
I can't answer your question, Karen, but I can answer the question(s) that popped into my mind when I read your title and content (so pedantically am guessing others are asking themselves something similar)!
1) Why is Karen doing Endevor Development using Agile?
She's not. Her developers are using "agile" for their application, are using z/dt as a development environment, and need access to the processes/processors Endevor performs.
2) What the heck is z/dt????
IBM Knowledge Center
"IBM® z Systems Development and Test Environment runs an IBM z Systems z/OS® distribution on a personal computer or workstation Linux environment. It creates an environment for mainframe application demonstration, development, testing, and employee education without the need for z Systems mainframe hardware. It enables z/OS, middleware, and other z/OS software to run on Intel and compatible computers. It is based on the IBM z Systems Personal Development Tool (zPDT®), which is included with the z Systems Development and Test Environment to emulate z Systems architecture with virtual I/O and devices."
Think back to the old days when you ran multiple MVS images under VM. This is the same thing except on Linux (I think). You set up multiple linux servers, all cloned from a “Gold Copy” server, which if I understand the process, is a dump of multiple volumes of all needed pieces (Endevor, db2, ims, tso, etc) from the Z/OS Lpar. See: https://www.ibm.com/support/knowledgecenter/en/SSQ2R2_14.0.0/com.ibm.zdt.overview.gs.doc/topics/c_product_overview.html.
So I’m trying to help figure out which “piece” be it Endevor or op/sys related is missing that causes a processor generated on said z/dt image to break LMAP
Technical Support Sr. Specialist
Software Code Management Team
860-902-3064 (IP phone)
Upcoming PTO: Aug 31 – Sept 6
So sorry but I can't help with your issue (not anything we've tried) but I am interested in what you're doing. At our site the plan is to use zDT for test boxes. We've got no plans to use it for any actual development (currently we plan to deliver the outputs from the existing development Systems to the zDT boxes to test them).
As the general idea of zDT seems to be "spin up, break, spin down, try again" what's your site's plans for using them for development?
We figured out the LMAP issue late Wed. afternoon. Somehow the MVS team picked up a 2 year old copy of the Endevor source (macro) library. GPPROCSS was pulling in an old copy of $PRDDB. ALOC=LMAP was being translated to ALLOC1=LMAP and ALLOC1 was being kicked out as an invalid keyword in the assemble step ☹. 1) I didn’t catch the PROC RC =4, only “saw” the NDVR RC =0. 2) Don’t get me started on why an Endevor process we have no control over, using Endevor macros we have no control over, lets an incorrect macro go through and not fail (fyi: $PRDDB is not footprinted). I am only involved in the Endevor admin part of z/dt, being pulled in by a DB2 variable in but from what I understand, you identify the minimum pieces and parts needed to support your lowest development stage (in our case UNIT), at regular intervals you dump all the volumes associated with these parts and restore them to a Z/dt image labeled the “Gold Copy” (don’t know if that’s a standard or just what we call it here). They then have to do all kinds of stuff on the system side to have Linux recognize these volumes and let it know it’s now an MVS image. Then you have x number of other Linux server for scrum work, that can get created or refreshed from at any time from the gold copy. For example, right now the Gold copy was rebuilt at the end of July, the DevOPs image is from May, and 2 other servers are from June, all working on different stories. They sue Topaz workbench and ftp to send stuff back and forth from “real life” and the gold copy. You log on to them just like you log onto regular TSO and to the naked eye it looks just regular tso. So on my 3270 emulator I have entries for TSO1, Gold, Scrum1 and Scrum2. I logon to them using my TSO1 credentials. In a nutshell that’s what I’ve learned in the last month, and they’ve been at this for quite a while now. If you need more detail than that, I can find out how much information they can share and if anyone more knowledgeable than myself is willing to meet with your people to describe the process in more detail.
Thanks Karen, I assume they're probably FTPing the source once they're happy with it. I'm just glad our development teams aren't up to the same tricks as it must cause headaches.
I’m not involved at that end, but I believe they are using Topaz workbench/Connect to add the stuff back into “real” Endevor. They haven’t even made it to the PDM discussion for merging all the different versions. Although I have never been a fan of sandboxes, I’m beginning to wish our Endevor was sandbox friendly. ☹