L
ON
W
ORKS
Router User’s Guide
2-7
As with configured routers, learning routers sometimes modify source addresses for
service pin messages to help prevent message loops.
Learning routers, in general, are less efficient in using channel bandwidth because they
always forward all messages with group destination addresses. Their advantage is
simplified installation since the installation tool does not need to know the network
topology to configure a learning router.
Message Buffers
As messages are received by a router, they are placed in an input buffer queue. By
default, this queue is limited to two message buffers to ensure that priority messages are
never queued behind more than one non-priority message. When forwarded to the
transmitting side of the router, priority messages have their own outgoing buffer queue.
This assures priority processing of these outgoing messages since the transmitting side
will send messages from the priority output buffer queue before sending messages from
the non-priority output buffer queue. Figure 2.4 summarizes the message flow through
the input and output buffer queues. This message flow is duplicated for messages moving
in the opposite direction, i.e., another set of input and output buffer queues exist for
messages flowing in the opposite direction.
.......
Input Buffer Queue
Direction of Message Flow
Priority Output Buffer Queue
Normal Output Buffer Queue
1
2
3
n-1
n
1
2
1
2
Message Receiving Side
Message Transmitting Side
Figure 2.4
Buffering Scheme for a L
ON
W
ORKS
Router
The size and count of the message buffers is limited by the amount of RAM on the
router. Each router side has 1,254 bytes of buffer space available. By default, this
space is allocated as two input buffers, two priority output buffers, and 15 non-priority
buffers. The default buffers are all 66 bytes, so the total RAM usage for the default
buffers is:
Summary of Contents for LONWORKS
Page 34: ......