IBM Mainframe Forum

Hi,
while creating the esds file i am getting this error

and the abend details

please tell me how to slove this problem.

thanks all

did You care to look at the frigging manual searching with IDC3009I 208

start from ...
http://www-03.ibm.com/systems/z/os/zos/bkserv/lookat/
enter the first message identifier that You wisely forgot to mention ... it should be IGD17103I
it will take You here
http://publibz.boulder.ibm.com/cgi-bin/ ... SPTM012041
where it will advise to search for the corresponding IDC3009I message

going back to
http://www-03.ibm.com/systems/z/os/zos/bkserv/lookat/
and using IDC3009I
it will take You here
http://publibz.boulder.ibm.com/cgi-bin/ ... SPTBB3009I
where You can follow the advice on searching or You can simply use Your browser search to <scroll> to the explanation for the 208 return code

but Yor system is behaving oddly
You should have received the message


the best place to seek for help is Your support

also another point ...
do not <type> the <exec> info
do a text cut and paste from Your 3270 session

IMBED and REPLICATE are interesting choices for an ESDS also. I suggest you look those up also

did You check the manuals or test before posting
http://publibz.boulder.ibm.com/cgi-bin/ ... T#FIRSTHIT

I did!

see also for historical reference ...
http://www-03.ibm.com/systems/z/os/zos/zos_sods.html

I consider myself suitably admonished. I have written 1000 times, RTFM before posting on here as they enjoy eating incorrect posts for breakfast.

I maintain though, better not to code them on an ESDS, removes any hint of confusion. Although, as was always the case when newbies create VSAM clusters.

'Here is a define i can copy, ooooo what are all those parameters i dont uderstand, what should i do?, i know i will leave them in as they obviously worked for the other guy so they must surely work for me'

NAHHHH...
and as I say often ... no need to stand corrected, better to sit corrected, it is more comfortable
my previous reply ( as worded) was a finger check (I was replying in parallel to two different posts )
a sentence was dropped, and it was too late to add it
and my true intention was just to point out that IMBED/REPLICATE were not related to the symptom described

'Here is a define i can copy, ooooo what are all those parameters i dont uderstand, what should i do?, i know i will leave them in as they obviously worked for the other guy so they must surely work for me'

yep... that' s the common approach ( not bad in se), but You forgot to add ...

and if something does not work instead of asking the guys whose snippets I used, I will ask on a forum,
where <dumb> people who know nothing about my organization practices will happily and snappily provide a solution
based on crucial info and requirements that I was very careful not to share

but given, the IT culture trend, software reuse, agile technology, sheer incompetence, time issues, ...
it is quite economical to reuse something that seems to work even if it is not according to the best practices

note on the documentation and implementation
the IMBED and REPLICATE things probably are part of the faulty logic in parameter parsing when it was changed to of ignore/not support them any longer

the equivalent old manual ( os/390 2.10 ) says
for imbed

IMBED|NOIMBED
Attention: IMBED is not supported; if specified, it is ignored and no message is issued.

but then it gives a lengthy explanation of how dasd is used

for replicate

REPLICATE|NOREPLICATE
identifies how many times each index record is to be written on a track.
This parameter applies only to key-sequenced clusters and variable-length relative record clusters.
REPLICATE
writes each index record on a track as many times as it will fit. With REPLICATE, rotational delay is reduced and performance is improved. However, the cluster's index usually requires more DASD space.
This parameter applies only to key-sequenced clusters and variable-length relative record clusters.

the newer manuals simply say

If you specify IMBED, KEYRANGE, ORDERED, or REPLICATE it will be ignored.

and as usual the TS has disappeared

follow on...
I forgot to add that single level dataset names are the worst of the worst

Enrico i feel your hurt, i do chuckle at some of the posts, they do seem to show not so much just a little knolwedge, but a TOTAL non understanding. It is interesting to let some of these loose on facilities costing millions, where they could have catastrophic implications for the companies concerned, with seemingly NO formal education in how to use this stuff.

Smacks of outsourcing and reducton of staff costs, but is soooooooooooo short sighted