Post by Kevin MonceauxPost by halfmegDL/I is included with the TK3 Update coming soon. It needs
someone to take it a little further to get it working properly.
Cool!!
Post by halfmegMy understanding is it works on MVT, but MVS has slightly
different MACROS and appendages (?) so it needs further work.
Does that mean the source was recovered? If I remember correctly
that was part of the problem getting it working under MVS. The
source was truncated due to a tape I/O error.
The tape never resurfaced to my knowledge, however whatever was not recovered didn't seem to prohibit DL/I from working on MVT.
Thought I had posted this previously but can't find it now.
---------------------------------------------------------
Volker posted in hercules-oc360 on Feb 13, 2005:
I have been working on getting DLI/360 to run on MVS. It did not
work. Then I tried to get a reference installation on MVT. It did
work (Thanks Kevin for the sample programs).
Taking these sample programs and the setup to MVS, I restarted my
efforts. Now, instead of getting an undocumented status code I got a
real MVS error message: DD Statement missing for (garbage)
Ooops - my porous memory suddenly remembered that a DBD contaisn an
OS DCB. Could it be that, by chance, the MVT DCB and the MVS DCB are
different? You bet. The MVS DCB has a 16 bytes prefix, which is
not there in MVT
I created an IDCB macro with the MVT DCB layout, and modified DBDGEN
macro to call IDCB instead of DCB - Voila, open now fails with a
"data mangement error" status code and an MVS message. Turns out
that we need an IEAAPP00 member which names the OSAM channel end
appendage. One IPL + CLPA later the open succeeds and the database
can be loaded.
But that's about it. There are more control blocks involved, like
IOB, DEB, also some EXCP programming . Any other process option but
load will fail with messages
IEC223I IFG0200V,UP0712,G,TESTDBO,13B,DLI001,DLI.TESTDB.OSAM
IEC999I IFG0TC0A,IFG0TC0B,UP0712 ,G ,DEB ADDR = 9ABF9C
when the OSAM part of the database is accesses. No reason code.
Looking at the job log I find a return code of 5728 (?? I always
knew that a RC cannot be more than 4095 ??)
I have a dump and a GTF trace for the problem. I have very little
spare time in the weeks to come. Now my questions:
1.) Where do I find documentation on how to write a channel end
appendage IGG019Z9
2.) Where do I find documentation on how to write a type 2 SVC for
MVT? for MVS?
3.) Is anybody willing to take a look at trace and dump at
http://www.bsp-gmbh.com/pub/dlidump.zip (340230 bytes) for me?
---------------------------------------------------------------
The link for the above trace and dump is gone now, perhaps due to a machine failure he suffered last year in which some data was not recovered IIRC.
To my knowledge there is little to no documentation on DL/I but IMS is the offspring so there may be some guidance in older IMS manuals.
Phil