FieldServer FS-8700-73 - Manual (Page 37)

FS-8700-73 BACnet MSTP Driver Manual Page 37 of 48
FieldServer Technologies 1991 Tarob Court Milpitas, California 95035 USA Web: www.FieldServer.com
Tel: (408) 262-2299 Fax: (408) 262-2269 Toll Free: (888) 509-1970 email: [email protected]
Appendix B.6. Rediscovering Offline Devices
The BACnet driver handles APDU retries internally and uses its own parameters, not the kernel parameters. The
kernel timeout and retry values should be left to default.
The following parameters can be configured on the Client Node:
APDU_Timeout - default value is 10s
APDU_Retries - default value is 3
The FieldServer will only send requests to the remote device once it has discovered it using the Who-Is / I-Am
process.
If a device does not respond and the APDU_Retries have been used up, the driver will revert to trying to discover
the device using Who-Is requests.
Note that there is also a background process of rediscovering devices independently of the polling process. This
occurs every 10 minutes.
Appendix B.7. Timeout Errors
NL TX Abort errors mean that the BACnet network layer was unable to transmit queued messages in time and gave
up on them. Outgoing packets are programmed with a 10 second timeout. . If packets have not been transmitted
within this 10 second window, they are discarded and an error is recorded.
Timeout errors may be experienced on start-up if the network is not yet fully synchronised or if there are a large
number of nodes on the network. The FieldServer can only transmit the I-Ams a few at a time, and only when it
receives the MSTP token.
It might help to give the MSTP network some time to settle before attempting a discovery.
It may be advisable to discover the Nodes in a number of smaller Node ID ranges (say 10 at a time). The Node_ID’s
can be established from the configuration.
Appendix B.8. PDU limit reached errors
A PDU limit reached error is returned when a device runs out of buffers for storing incoming or outgoing messages
(PDU = protocol data unit). A possible scenario is when a large number of server nodes is configured and one or
more global Who-Is requests is received. The FieldServer will then need to transmit a response for each one, but if
the MSTP network is busy it might not get enough airtime to send them all out. Another possibility is a buildup of
unsent COV or Event Notifications.