Browse Prior Art Database

IMS/ESA MSC/VTAM Bandwidth Improvement

IP.com Disclosure Number: IPCOM000121361D
Original Publication Date: 1991-Aug-01
Included in the Prior Art Database: 2005-Apr-03
Document File: 2 page(s) / 79K

Publishing Venue

IBM

Related People

Georgiadis, L: AUTHOR [+4]

Abstract

A program is disclosed that implements a new MSC (Multiple Systems Communication)/VTAM communication algorithm to the IMS/ESA product. It allows multiple messages in a group to be transmitted and waits for only one Definite Response (DR) from the receiver. In addition, it reduces the number of Check Writes (CHKW) by the receiver from one per message to just one per group. The overall reduction in I/O time and network traffic are tremendous. Session Initiation

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

IMS/ESA MSC/VTAM Bandwidth Improvement

      A program is disclosed that implements a new MSC
(Multiple Systems Communication)/VTAM communication algorithm to the
IMS/ESA product.  It allows multiple messages in a group to be
transmitted and waits for only one Definite Response (DR) from the
receiver.  In addition, it reduces the number of Check Writes (CHKW)
by the receiver from one per message to just one per group.  The
overall reduction in I/O time and network traffic are tremendous.
Session Initiation

      As a result of entering /RST specifying threshold values, a
"send table" will be created on both sides of the link after exchange
of restart record which contains threshold values.  Each row in the
send table will include the following information regarding a message
in a group:
      1.   DRRN (Disk Relative Record Number) - 4 bytes
      2.   Queue Number - 1 byte
      3.   Queue name - 8 bytes
           Queue Manager uses the above three fields to
           uniquely identify a message for dequeuing purpose
           after the sender receives a DR or an ER from the
           receiver.
      4.   Only in chain (OIC) or last in chain (LIC)
           sequence number - 2 bytes
           When an ER is received, the sender needs to check
           the sequence number from the Request Parameter
           List (RPL) and the send table to decide which
           messages can be dequeued.
      5.   Recoverable or nonrecoverable indicator - 1 bit Since
nonrecoverable messages will be tossed away by Queue Manager in
restart, this indicator is a safety check if dequeue does not give a
good return code back.
      6.   Return code - 2 bytes
           It contains the return code from Queue Manager
           dequeue call to allow some checking by the caller.

      After session initiation is complete, an X'6C' type log record
will be logged on both sides of the link with extra fields containing
threshold values.  This is needed for the purpose of session restart.
Normal Message Processing

      Before a mes...