42
MultiFRAD II User Guide
Frame Relay Setup
The Frame Relay dialog box displays the CIR Measurement Interval in milliseconds, the
Management Type and details of that management type, and the number of DLCI’s that are active.
The MultiFRAD can detect DLCIs and the Management Type when you are connected to an active
frame relay service. So it is important that if you change any frame relay parameters, that the new
parameters agree with the way your frame relay service is provisioned.
In the Frame Relay Setup dialog box, you can change the Management Type by choosing one of the
three options and then the detail parameters for that option are displayed to the right of the
Management Type. For example, in the above dialog box, Annex D is the chosen management type
with the Full Status Enquiry Interval [N391] set to 6 down thru Polling Verification Timer [T392] set to
15.
Usually, you will have no need to change the management type because the MultiFRAD will
automatically detect the network’s management type at power-up, or when the network becomes
active. This auto-detection feature takes about 1 to 2 minutes to detect the management type and get
a list of the DLCI’s.
Occasionally the Management Type parameters can be changed to correct a problem. In these
cases, making the following changes could alleviate the problem:
1. Change the Link Integrity Verification Timer. This feature sets how often the MultiFRAD verifies
that the link is good. Reducing the setting from the default value of 10 seconds to a lower value; e.g.,
5 seconds, will increase the frequency of those verifications.
2. Change the Full Status Enquiry Interval to affect how often DCLI Status is updated. This
parameter defines how often the MultiFRAD will send a request for DLCI status information along
with the Link Integrity Verification request. The default value is to include such a request every sixth
time the link integrity is checked. A value less than every sixth request will enable faster detection of
active/inactive DLCIs, but will also produce a slight increase in management overhead.
3. Change the Monitored Events Count and the Error Threshold to affect the MultiFRAD’s
sensitivity to errors on the link. Every management frame received by the MultiFRAD is considered
an event, and if the link integrity sequence numbers are wrong, for example, it is considered an
“errored” event. With the default settings, if Error Threshold is 3 and Monitored Events Count is 4,
and 3 out of 4 events are “errored”, then the link is considered to be bad, and the FRAD stops
sending on all DLCIs and restarts the initial procedures to activate the link.
4. Change the CIR Measurement Interval to change the time frame in which the FRAD calculates
throughput for each DLCI. By default, every 1000 milliseconds (1 second), the MultiFRAD calculates
how much information has been sent on each DLCI, and depending on the congestion management
option for a DLCI the MultiFRAD may stop sending on that DLCI because it has exceeded its CIR or
CIR+Be rating. By changing the CIR Measurement interval, the throughput calculations can be made
more or less often.
Note: performance could be compromised if the MultiFRAD is set to calculate throughput too often.