Browse Prior Art Database

Performance Improvements for Processing Frame Relay Packets

IP.com Disclosure Number: IPCOM000104435D
Original Publication Date: 1993-Apr-01
Included in the Prior Art Database: 2005-Mar-19
Document File: 2 page(s) / 49K

Publishing Venue

IBM

Related People

Ashton, J: AUTHOR [+4]

Abstract

Disclosed is a method of operation that allows efficient placement of data in buffers, received from various sources across a packet switched network. This allows efficient connections to multi-protocol routers (i.e., IBM 6611) and their downstream Token Ring devices, to the IBM 3745 and Network Control Program (NCP).

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 96% of the total text.

Performance Improvements for Processing Frame Relay Packets

      Disclosed is a method of operation that allows efficient
placement of data in buffers, received from various sources across a
packet switched network.  This allows efficient connections to
multi-protocol routers (i.e., IBM 6611) and their downstream Token
Ring devices, to the IBM 3745 and Network Control Program (NCP).

      Frame Relay Data Link Connection identifiers (DLCI) are defined
with values ranging from 16 to 1007.  Constructing a table of 128
bytes provides enough bits to define a Bit Mask for each DLCI.  The
bit can be used to select one of two offsets in the receive buffers.
The DLCIs are defined to NCP by the customer, and used by the Network
Definition Facility (NDF) to generate the Bit Map table.

      At scanner (High Performance Transmission SubSystem - HPTSS)
initialization time, the address of the Bit Map table is passed to
the scanner.  The table is transferred to scanner storage by the
scanner hardware.  The offsets to be used by the scanner to place
data in the receive buffers is also passed at initialization time.
Figure 1 shows an example of offsets set to x'1A' and x'2E'.

      As each frame is received (1), the DLCI (2) is examined.  The
DLCI value is used to calculate the correct bit offset in the Bit Map
Table (3).  If the corresponding bit is zero, the first defined
offset is used to control placement of data in the NCP buffers (4).
If the corresponding b...