|
Ricardo,
>> I am developing a multiprocess client/server program for connecting tcp/ip
>> clients to a server process on an OSF host. The system uses the EPI client to
>> connect to the CICS 2.1 system running on OSF 3.2C, with Oracle 7.1.6 as the
>> database engine. The application is a migration from IBM MVS.
>>
>> On some transactions, when they return an error condition in the application,
>> Abend Code ATDD, the EPI client receives some unexpected codes from the host.
An ATDD abend occurs if you attempt to access a transient data queue that
is disabled. Is this the problem?
>> May be, there is some relationship between the transaction not executing
>> a syncpoint before ending the transaction on the OSF host and this problem,
>> because some transactions modified to do explicit syncs do not exhibit this
>> behavior any longer. Is this a right path to fix this problem ?
What do you see as the problem? The ATDD abend, or the datastream which
CICS is generating to send to the epi client when the error occurs? (or
of maybe both). If both can I suggest that you look at solving them
separately.
If you are concerned about the 3270 datastreams produced by CICS, can
you check what is produced if you try running the transaction from a
cicsterm?
Please let me know if this helps,
Regards,
Helen.
|
| Hello, Helen.
I am interested only in the 3270 data stream received in the EPI client.
Mainly
I wonder if this "unexpected codes" received previous to the abend message are
a
normal CICS host reply in case of abend or is related to the type of terminal
(OSF EPI client) where the transaction is executed.
I talk about "unexpected" because I detected multiple EventSend events in the
EPI
getevent function and I expect receiving only one EventSend per transaction.
Is
this a normal EPI behavior and what is the meaning of this codes?
I cannot post now the 3270 data stream received when launching the transaction
from a cicsterm. I have to visit the customer site to get it. By the way, how
can
I capture the data stream, cicsterm does not have a trace utility or does it?
Thanks for your quick answer,
Ricardo Lopez Cencerrado
[Posted by WWW Notes gateway]
|
|
Hi Ricardo,
>>Mainly I wonder if this "unexpected codes" received previous to the abend
>>message are a normal CICS host reply in case of abend or is related to the
>>type of terminal (OSF EPI client) where the transaction is executed.
The "unexpected codes" are valid 3270 datastreams, although
they don't make a lot of sense. I suspect that they are are generated
by the transaction before the abend failure. Does the transaction attempt
to send anything back to the client? If the transaction attempts to send
information back to the client, it is possible that CICS will modify the
data dependent upon the terminal type and what it supports - I have not
seen this cause a problem.
>>I talk about "unexpected" because I detected multiple EventSend events in the
>>EPI getevent function and I expect receiving only one EventSend per
>>transaction.
This depends upon the way in which the transaction is written (how
many maps it sends back to the client for example).
>>I cannot post now the 3270 data stream received when launching the transaction
>>from a cicsterm. I have to visit the customer site to get it. By the way, how
>>can I capture the data stream, cicsterm does not have a trace utility or
>>does it?
I can't think of a way to capture the data stream between a region and
cicsterm, however given what is being sent back to your Epi application,
you should be able to see if similar datastreams are being sent to a
cicsterm.
>>Is this a normal EPI behavior and what is the meaning of this codes?
The following is my enterpretation of the 3270 datastreams in ASCII which
CICS is generating, please note that I'm not very experienced in intrepreting
so if I've got them wrong, someone please correct me ;-)
Hope this helps,
Helen.
------------------------------------------------------------------------------
I have preceded my interpretations with hp>
Note that the data values are in ASCI. All but the last data stream
have 19 at the end. In my book 19 maps to the EM interface code which
I'm not familar with. However in the Epi3270 server shipped with CICS
19 is mapped to an the '!' character. Can someone help here?
---- **** ----
Session Transcript
----
DebugOn value is TRUE
DebugVerbose value is TRUE
DebugDir value is /usr/users/ricardo/logdir
Log file /usr/users/ricardo/logdir/17299_#23_01_1997#20_49.log created
EPI initialized on daemon
The client inet address is: 16.190.192.70
The client port is: 15108
RegionName: TCICSORA Timeout: -1 TermType: hft BufSizeIn: 8192 BufSizeOut:
8192
AddTerminal command sent to host OK
Session Terminal LICO7299
Pollin detected on poll
Pollnorm detected on poll
Transaction submitted to host
The following message was received from CICS:
<<<
112020131120202037353111204419>>>
hp> 11 - Set buffer address
hp> 20 20 - Location row 1, column 1
hp> 13 - Insert cursor
hp> 11 - Set buffer address
hp> 20 20 - Location row 1, column 1
hp> 20 37 35 31 - <space> 7 5 1
hp> 11 - Set buffer address
hp> 20 44 - Location row 1, column 5
hp> 19 - see above note.
Eventt>>Receive data event detected on thread
The following message was received from CICS:
<<<
11202013112020202F3211204319>>>
hp> 11 - Set buffer address
hp> 20 20 - Location row 1, column 1
hp> 13 - Insert cursor
hp> 11 - Set buffer address
hp> 20 20 - Location row 1, column 1
hp> 20 2F 32 - <space> / 2
hp> 11 - Set buffer address
hp> 20 43 - Location row 1 column 4
hp> 19 - see above note.
Eventt>>Receive data event detected on thread
The following message was received from CICS:
<<<
112020131120202030303030303011204719>>>
hp> 11 - Set buffer address
hp> 20 20 - Location row 1, column 1
hp> 13 - Insert cursor
hp> 11 - Set buffer address
hp> 20 20 - Location row 1, column 1
hp> 30 30 30 30 30 30 - 0 0 0 0 0 0
hp> 11 - Set buffer address
hp> 20 47 - Location row 1, column 8
hp> 19 - see above note.
Eventt>>Receive data event detected on thread
The following message was received from CICS:
<<<
11202013112020203311204219>>>
hp> 11 - Set buffer address
hp> 20 20 - Location row 1, column 1
hp> 13 - Insert cursor
hp> 11 - Set buffer address
hp> 20 20 - Location row 1, column 1
hp> 20 33 - <space> 3
hp> 11 - Set buffer address
hp> 20 42 - Location row 1, column 3
hp> 19 - see above note.
Eventt>>Receive data event detected on thread
The following message was received from CICS:
<<<
1120201311202020202011204319>>>
hp>11 - Set buffer address
hp> 20 20 - Location row 1, column 1
hp> 13 - Insert cursor
hp> 11 - Set buffer address
hp> 20 20 - Location row 1, column 1
hp> 20 20 20 - <space> <space> <space>
hp> 11 - Set buffer address
hp> 20 43 - Location row 1, column 4
hp> 19 - see above note.
Eventt>>Receive data event detected on thread
The following message was received from CICS:
<<<
112A3014292220112A301024232C05945525A
303134303136453A205472616E73616374696F6E
2027444F3039272C204162656E64202741544444
272C20617420274C493939272E1129221D20>>>
hp> 11 - Set buffer address
hp> 2A 30 - Location row 24, column 1
hp> 14 - Repeat to address
hp> 29 22 - Location row 24, column 80
hp> 20 - <space>
hp> 11 - Set buffer address
hp> 2A 30 - Location row 24, column 1
hp> 10 24 23 2C 05 9 ? $ # , ? - Note can someone clarify
hp> 45 52 5A 30 31 34 30 31 36 45 3A E R Z 0 1 4 0 1 6 E :
hp> 20 54 72 61 6E 73 61 63 74 69 6F 6E <space> T r a n s a c t i o n
hp> 20 27 44 4F 30 39 27 2C 20 41 62 65 <space> ' D O 0 9 ' , <space> A b e
hp> 6E 64 20 27 41 54 44 44 27 2C 20 61 n d <space> ' A T D D ' , <space> a
hp> 74 20 27 4C 49 39 39 27 2E t <space> ' L I 9 9 ' .
hp> 11 - Set buffer address
hp> 29 22 - Location row 24, column 80
hp> 1D - Start field
hp> 20 - Protected
Eventt>>Receive data event detected on thread
Eventt>> Transaction end event detected by thread
Eventt>> Transaction Abend Code: ATDD
[Posted by WWW Notes gateway]
|