Re: Async comm not timing out

[Mark Lazarus <mlazarus@xxxxxxxx>]

Buck,

>>>This week's question is for the communications programmers.
>>>I have an async program that does an invite, then a read.
>>>The WAITRCD is set to 15 seconds, and works most of the 
>>>time.  The situation where it doesn't work is when I 
>>>have lots of parity errors on the line.  What happens is...
>>>nothing.  The line, controller and device all show ACTIVE
>>>but the READ doesn't time out.  If I cancel the job, the
>>>controller and device go to FAILED.  If I vary the line, 
>>>ect. off and on and re-start the job, everything works fine 
>>>until the parity errors pile up again.
>>
>> I'm going to guess at what is happening here.  You have a *PSSR routine to
>>do processing for the timeout condition.  When getting those line errors,
>>the program goes into the *PSSR.  What is probably going on is that the
>>timeout occurs during error processing, creating an error upon error
>>condition.  
>>
>> I think the solution is to explicitly test for those error conditions in
>>your subroutine.  Maybe you should post the pgm code?
>
>Yes, there's a *PSSR, but I'm not using it in this case: I have indicators
>on EOF and ERROR.  What happens is the line is dropping and the READ
>is never satisfied.  By that, I mean that the code sits at ICFW (the only
READ
>in the program!) and never executes the tests for MAJ/MIN, EOF or ERROR.
>
>Here's the code snippet:
>     C                     SETOF                     80               
>     C           *IN80     DOWEQ*OFF                                  
>     C                     WRITEINVITE                 81             
>     C                     READ DMSMTXFM               8180           
>     C   80                LEAVE                                      
>     C* Timed out; drop out                                           
>     C           MAJCOD    IFEQ '03'                                  
>     C                     MOVEL'Y'       PMTERR                      
>     C                     LEAVE                                      
>     C                     ENDIF                                      
>     C* Comm error; drop out                                          
>     C           *IN81     IFEQ '1'                                   
>     C                     MOVEL'Y'       ERRFLG                      
>     C                     LEAVE                                      
>     C                     ENDIF                                      
>     C* Process result
>
>81 is ERROR, 80 is EOF.
>
>I am wondering if the line wierdness is generating XOFF characters
>that are hosing up the READ...

 I think you s/b using the *PSSR and test for the maj/min codes in there.
Also, use the factor 2 function on the ENDSR to control the logic.

 -mark

+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to "MIDRANGE-L@midrange.com".
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---