Configuring Resource Availability Indicator
The Resource Availability Indication (RAI) is a mechanism by which gateways can report the status of resources to the gatekeeper. If RAI is enabled, the gateway tracks the status of DS0 channels and DSP resources. This feature only tracks DS0s and DSP resources; analog trunk ports are not included.
RAI is useful in locations where more than one gateway is servicing a destination. An example of this might be a large location with multiple PSTN connections. You can have multiple trunks dispersed across several gateways for increased capacity and redundancy.
When RAI is enabled, resources are tracked against a defined threshold. When the high resource threshold is exceeded, an RAI message is sent to the gatekeeper indicating that the gateway is almost out of resources. When the resources have become available again and the low threshold is met, another RAI message is sent to the gatekeeper indicating that the gateway is no longer resource constrained.
The gatekeeper assigns a priority to gateways. If multiple gateways service the same destination, the gatekeeper routes the calls based on the gateway priority. If all the gateways have the same priority, the gatekeeper load-balances across them. The default gateway priority is 5. You can change the priority of a specific gateway using the zone prefix gatekeeper-name e164-prefix [gw-priority priority gw-alias [gw-alias,...]] command. You can set the priority in a range from 0 to 10. The gw-alias is the H323-ID of the gateway that you want to change.
When the gatekeeper receives an RAI from a gateway indicating that it is almost out of resources, it lowers the priority of that gateway to 1. When that occurs, the gatekeeper sends calls to the other gateways that are servicing the destination with higher priorities. If no other gateway has a higher priority or if all gateways in that zone have priority 1, the gatekeeper still sends calls to the gateway that sent the RAI message.
When the low threshold is met and the gatekeeper receives the RAI message indicating that resources are again available, it restores the priority of the gateway to the original value.
No configuration is necessary on the gatekeeper to enable RAI.
RAI is enabled on the gateway using the resource threshold [high percentage-value] [low percentage-value] [report-policy {idle-only | addressable}] command in gateway configuration mode. The optional report-policy parameter determines how to calculate resource utilization:
- idle-only Utilization = (in-use)/(in-use plus free)
- addressable Utilization = (in-use plus disabled)/addressable
The addressable channel calculation includes disabled channels; therefore, it provides a more accurate percentage of available channels.
You can also optionally code the lrq reject-resource-low command on the gatekeeper. Normally, all requests are processed even if the destination gateways are reporting that they are almost out of resources. If the lrq reject-resource-low command is included, the gatekeeper can reject (issue a Location Reject, or LRJ) requests from remote zones to a destination that is resource constrained.
Example 17-19 shows the configuration that is necessary to enable RAI on a gateway.
Leeds#show running-config Building configuration... ! ! Unnecessary output deleted ! controller E1 1/0 ds0-group 1 timeslots 1-4 type r2-digital r2-compelled ani ! ! interface Serial0/0.102 point-to-point ip address 10.13.1.2 255.255.255.0 frame-relay interface-dlci 102 h323-gateway voip interface h323-gateway voip id leeds multicast h323-gateway voip h323-id leeds@cisco.com h323-gateway voip tech-prefix 1# h323-gateway voip bind srcaddr 10.13.1.2 ! ! voice-port 1/0:1 ! voice-port 2/0/0 ! voice-port 2/0/1 ! voice-port 2/0/2 ! voice-port 2/0/3 ! dial-peer voice 99 voip destination-pattern .T session target ras codec g711ulaw ! dial-peer voice 1 pots destination-pattern 0113496.... no digit-strip port 1/0:1 ! ! gateway resource threshold high 75 low 50 ! end |
In the example, only four timeslots are enabled on the E1 trunk. The high threshold is 75 percent, and the low threshold is 50 percent. Voice port 1/0:1 is referenced in a dial peer, making the channels addressable. This is importantonly DS0s that are associated with a POTS dial peer are monitored. Usage of available DSP resources is also monitored and compared to the thresholds for RAI reporting.
Even though the gateway has analog voice ports, only the DS0 ports are included in calculating resource usage.
You can see which resources are being monitored and which thresholds are in use by using the show gateway command, as demonstrated in Example 17-20.
Leeds#show gateway H.323 ITU-T Version: 4.0 H323 Stack Version: 0.1 H.323 service is up Gateway leeds@cisco.com is registered to Gatekeeper leeds Alias list (CLI configured) H323-ID leeds@cisco.com Alias list (last RCF) H323-ID leeds@cisco.com H323 resource thresholding is Enabled and Active H323 resource threshold values: DSP: Low threshold 50, High threshold 75 DS0: Low threshold 50, High threshold 75 |
To see the status of monitored DS0 and DSP resources, use the show call resource voice stats command. To see whether any of the configured thresholds have been exceeded, use the show voice resource call threshold command. Example 17-21 shows the results of both of these commands, in addition to the show gatekeeper endpoint command.
[View full width] CURRENT STATUS OF MONITORED RESOURCES Leeds#show call resource voice stats Resource Monitor - Dial-up Resource Statistics Information: DSP Statistics: Utilization: 0.00 percent Total channels: 76 Inuse channels: 0 Disabled channels: 0 Pending channels: 0 Free channels: 76 DS0 Statistics: Utilization: 0.00 percent Total channels: 31 Addressable channels: 4 Inuse channels: 0 Disabled channels: 4 |
In the example, you can see that 76 DSP resources exist, and all are free. Four DS0 resources are available, and all four are disabled, leaving zero free. This means that the gateway should be telling the gatekeeper that it is resource constrained via an RAI message. The show gatekeeper endpoints command in the example lists the leeds gateway with an O flag. This confirms that the gatekeeper has received the RAI message, and the gateway is flagged as out of resources.
You can also see the actual RAI message sent by the gateway and the gatekeeper response using the debug h225 asn1 trace command. Example 17-22 shows the RAI that is sent to the gatekeeper.
[View full width] Leeds#debug h225 asn1 *Jun 25 19:41:27.537: RAS OUTGOING PDU ::= value RasMessage ::= resourcesAvailableIndicate : |
In the trace, the almostOutOfResources flag is set to TRUE, informing the gatekeeper of a resource constraint. The gatekeeper acknowledges with a resourcesAvailableConfirm message, changes the priority of the gateway to 1, and sets the O status flag.