Re:Re: [IDMSVENDOR-L] Strange thing in IDMS 17.0 SP1 DCLOG

[ca.portal.admin]

are you checking for ""both"" log full msgs?

DC050001
DC050004

i believe DC050004 is spit out when the log is 100% full - this message is
NOT checked by vanilla WTOexit

i have seen a log fill up so quicly due to a dump that the DC050001 is
never issues - only the DC050004 - and the vamilla wtoexit would never
catch it




Chris Hoelscher
IDMS/DB2 Database Architect
Humana Inc
502-476-2538
[e-mail]

you only need to test the programs that you want to work correctly






From:
""William M. Allen, Jr."" <[e-mail]>
To:
[e-mail]
Date:
05/05/2010 04:29 PM
Subject:
[IDMSVENDOR-L] Strange thing in IDMS 17.0 SP1 DCLOG
Sent by:
IDMS 3rd-party providers forum <[e-mail]>



Hello All:



I just upgraded another Client to 17.0 SP1 and we are having a few issues;
one was with TSSMAI from Top Secret and there is a fix for that from CA.



The next issue is that the CV is not spitting out the log % full message
so
what is happening is that the log is filling up and the offload job is
never
getting submitted?



The exit is fine because it is submitting the Journal archive just fine.



The Messages for the Journal and log have the same destination and OS
route
codes, I have checked all of that. I have even restored and re-received
and
applied the startup user modification with no change.



The Log was formatted before we brought the CV up using IDMS 17.0 SP1
software.



We also had an issue with easy test causing system module program checks
near RHDCUXIT with all the tool kit exits installed. We have pulled all
the
exits from RHDCUXIT and easy test works fine. We still have to work with
the
vendors on a resolution for this one.



The last issue is the submission of the log archive, I know the exit is
working and the log has filled many times, but the IDMS Region is simply
not
issuing the log % full message?



And here is another strange thing when we view the log with LOGD we see
these system module program checks from four days ago and the log has
filled
at least five times since then because we have to archive it manually.



So my question is how does IDMS issue that log message, is there some sort
of header record in the log that might be corrupted?



We can always format the log but the Client is reluctant to do that
without
an explanation of why this is happening.



William M. Allen, Jr.

ARCH Consulting Associates, Ltd.

[phone]



The information transmitted is intended only for the person or entity to
which it is addressed and may contain CONFIDENTIAL material. If you
receive
this material/information in error, please contact the sender and delete
or
destroy the material/information.



The information transmitted is intended only for the person or entity to which it is addressed and may contain CONFIDENTIAL material. If you receive this material/information in error, please contact the sender and delete or destroy the material/information.
"
IDMS Public Discussion Forum
[e-mail]
SMTP
[e-mail]
[e-mail]
SMTP








Normal

Normal
"Re: question about ""get storage"""
"OK, you caught me. I had simplified the situation so that my problem would=
not get lost in the details.

PROGRAMB is not actually misbehaving. We have a situation here where user =
programs pass XML to and from some common routines. The size of the XML ca=
n vary from a few hundred bytes to infinitely large. =20

The common routines are successfully passing all different lengths of stora=
ge to and from each other, using a clever solution provided by Chuck Hardee=
on this list some time ago. =20

The user programs also work fine when they use Chuck's solution. However, =
this solution is a bit tricky, and I was trying to simplify things for ""Joe=
the programmer"". I thought that he could code the storage area in the lin=
kage section large enough for the maximum expected XML, and that the ""TO"" o=
ption would prevent overflow if the XML was larger than expected, thus prot=
ecting the program from itself, and not requiring Joe to have to write a lo=
t of length-checking code.

But the ""TO"" option does not provide this protection, so I need to reconsid=
er my recommendation for the user programs. I think I will look into the p=
ossibility of encapsulating Chuck's solution in copybooks. This has worked=
well here for other types of ""tricky"" programming such as SQL table proced=
ures.