asynch IDLTMR parameter

SC,
The AS/400 help for the IDLTMR says:

 Specifies the time (in 0.5 second intervals) that the system waits
 between characters before the adapter forwards the receive buffer to
 the system.

 Note:  Idle timer is also referred to as inter-character timer.

My guess is that the remote side is leaving a gap of more than 0.5
seconds between certain characters.  You said a datascope trace showed
the characters were being sent contiguously.  It's been awhile since I've
used a datascope, but does the datascope show the actual time delays
between characters on the asynch line?

The fact that increasing the IDLTMR parameter from 0.5 seconds to 2.0
seconds fixed your problem leads me to believe that the remote end is,
indeed, leaving pauses between some asynch characters.

Marty Keech  IBM AS/400 APPC, ICF, CPI-C (and a little asynch)


Hi,
>
>We are running on V3R6 on a PowerPC 530-2153 box, recently one of our
>online applications encountered a problem on its ASYNC communication line
>which caused the driver program to abend. There was no link interruption
>at that time and the ASYNC line, controller remained in ACTIVE status while
>the device was in VARIED ON status - as the driver program had abended. We
>managed to capture a communication trace using SST and found that messages
>were split into two parts when there were received at the AS/400 though the
>datascope trace (taken from a data analyzer connected on the AS/400 ASYNC
>line) showed that continuous full messages were being sent by the
>originating Host at the other end. This is acceptable as long as the two
>blocks are sent by the AS/400 to the application within a time limit - we
>suspect it is controlled by the IDLTMR (specified in the AS/400 ASYNC line
>description). However, there were instances where the second block was
>received after the IDLTMR had expired and as a result, a single continuous
>message was processed as two separate incomplete messages - causing our
>application driver to abend.
>
>The IDLTMR was changed from 1 (or 0.5 seconds) to 4 (or 2 seconds) which
>seem to improve the situation as the link became more stable. However, we
>are unsure whether the IDLTMR is the actual cause of the problem and would
>appreciate if anyone out there can advise the following :
>
>1. Why does the AS/400 spilt each incoming message into two parts although
>   it was sent as one continuous datastream without any break in
>   transmission?
>2. What is the actual function of the IDLTMR?
>3. From the enclosed comm trace, how do we interprete the Record Timer?
>   Is it in seconds, milliseconds and is this related to the IDLTM setting?
>
>
>Thank You.
>
>SC Loo
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* 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 MAJORDOMO@midrange.com and specify            *
* 'unsubscribe MIDRANGE-L' in the body of your message.  Questions      *
* should be directed to the list owner / operator: david@midrange.com   *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *