|
|
Cisco Resource Pool Manager Server (Cisco RPMS) enables telephone companies and Internet service providers (ISPs) to count, control, manage, and provides accounting data for shared resources for wholesale Virtual Private Dial-up Network (VPDN) and retail dial network services across one or more network access server (NAS) stacks. Cisco RPMS offers a web browser-based configuration utility that enables you to:
This chapter describes concepts you should understand before you configure or plan to configure Cisco Resource Pool Manager Server (Cisco RPMS).
Cisco RPMS differentiates dial customers through configured customer profiles based on the dialed number (DNIS) and call type determined at the time of an incoming call.
The incoming side of the customer profile determines if the call will be answered using parameters such as DNIS and call type from the assigned DNIS group and session limits. The call is then assigned the appropriate resource as defined in the resource groups. Each configured customer profile includes a maximum allowed session value and an overflow value. As sessions are started and ended, session counters are incremented and decremented so customers receive the appropriate call treatments.
The outgoing side of the customer profile directs the answered call to the appropriate destination:
Customer profiles contain the following components:
Figure 1-1 shows the major components of a customer profile.
There are three types of customer profiles in Cisco RPMS:
Configured DNIS groups and resource data can be assigned to customer profiles. These customer profiles are selected by the incoming call DNIS number and call type and then used to identify resource allocations based on the assigned resource groups and resource services.
After the call is answered, customer profiles can also be associated with VPDN groups so the configured VPDN sessions, home gateway information, and other data necessary to set up or reject a VPDN session are applied to the answered calls. VPDN group data includes associated domain name or DNIS, home gateway information, maximum MLP bundles per home gateway, and maximum links per MLP bundle.
Default customer profiles are used to provide session counting and resource assignment to incoming calls that do not match the configured DNIS groups. While specific resources and DNIS groups can be assigned to customer profiles, default customer profiles allow resource pooling for the calls that do not match the configured DNIS groups. Retail dial services and domain name-based VPDN services use default customer profiles.
You can configure multiple default customer profiles. When multiple default customer profiles are used, the call type (speech, digital, V.110, V.120) of the default DNIS group is used to identify which default customer profile to use for an incoming call. At most, four default profiles (one for each call type) can be configured on each Cisco RPMS server.
The backup customer profile can contain all of the elements defined in a standard customer profile, including base-size or overflow parameters. However, when the connection between the Cisco NAS and Cisco RPMS is unavailable, session counting and session limits are not applied to incoming calls. Also, after the connection is reestablished, there is no synchronization of call counters between the Cisco NAS and Cisco RPMS.
In the case of an 800, 888, or similar number, the DNIS number delivered to the NAS by the telco switch is different from the number dialed by the user. Cisco RPMS provides an optional reference number in each DNIS group for numbers that require public network database lookup and result in a POTS-routable DNIS.
Call types from calls originating from ISDN, SS7, and CAS (CT1, CT3, and CE1) are used to assign calls to the appropriate resource group. Call types for ISDN are based on Q.931 Bearer Capability. Call types for CAS are assigned based on static channel configuration or dynamically based on DNIS. Supported call types are:
Cisco RPMS enables you to maximize the use of available shared resources across multiple Cisco NASes for various resource allocation schemes. Cisco RPMS allows you to combine your Cisco NAS resource groups with call types (speech, digital, V.110, and V.120) and optional resource modem services into resource data assignments. Resource data can be configured for customer profiles and assigned to incoming calls through DNIS groups and call types.
Resource groups are configured on the Cisco NAS and assigned by Cisco RPMS using the customer profile. Resource groups represent groupings of similar hardware and/or services that are static and do not change on a per-call basis. Resource groups can define resources that are port-based or not port-based. Port-based resources are identified by physical location, such as a range of port/slot numbers (for example, modems or terminal adapters). Resources that are not port-based are identified by a single size parameter (for example, HDLC framers or V.120 terminal adapters).
Cisco RPMS resource assignments contain combinations of Cisco NAS resource groups, optional resource modem services, and call types. NAS resources in resource groups that have not been assigned to a Cisco RPMS customer profile will not be utilized.
In the NAS, use the show resource-pool resource command to display the configured NAS resource groups. Use the show resource-pool resource [name] to display the contents of a configured NAS resource group.
Services supported by a resource group are determined by the combination of hardware and portware installed. Services or specific functions can be configured and applied to a resource dynamically on a per-call basis. A customer profile can have an optional modem service assigned to it to be used on the allocated resource on a per-call basis.
Ensure the resource service commands and parameters are entered correctly. Cisco RPMS does not perform any validation of the resource service values. As a result, incorrect service commands and/or parameters might cause calls to fail.
VPDN groups contain DNIS-number or domain-name assignments so answered calls can be assigned the configured VPDN home gateway information and other data necessary to set up or reject the VPDN session. VPDN group data includes session limits, associated domain name or DNIS, home gateway information, maximum MLP bundles per VPDN group, and maximum links per MLP bundle.
For DNIS-based VPDN dial service, VPDN groups are assigned to customer profiles based on the incoming DNIS number and the configured DNIS groups for the initial resource allocation and customer session limits.
For domain-based VPDN dial service, VPDN groups are assigned to the default customer profile with the matching call type/default DNIS group assignment. The default customer profile contains the initial resource allocation and customer session limits.
Cisco RPMS allows you to specify multiple home gateway IP endpoints for a VPDN group. If three or more IP endpoints are specified, Cisco RPMS uses a load balancing method to ensure traffic is distributed across the IP endpoints. This load balancing method obtains the first three available IP endpoints as listed in the Valid IP Endpoints page and selects the first available IP endpoint as the one to be used. For the next request, the next three IP endpoints are obtained from the list, and the first available one used. Cisco RPMS also rotates the use of the available IP endpoints within the groups of three so that the second and third available IP endpoints will be used before the first IP endpoint is used a second time. For example:
Available IP Endpoints:
IP Endpoint1, IP Endpoint2, IP Endpoint3, IP Endpoint4, IP Endpoint5, IP Endpoint6
First Home Gateway Request:
IP Endpoint1, IP Endpoint2, IP Endpoint3
Selected: IP Endpoint1
Second Home Gateway Request:
IP Endpoint4, IP Endpoint5, IP Endpoint6
Selected: IP Endpoint4
Third Home Gateway Request:
IP Endpoint2, IP Endpoint3, IP Endpoint1
Selected: IP Endpoint2
Figure 1-2 shows the relationship of Cisco RPMS and RMP.
Cisco RPMS uses the following call processing to manage dial services.
Incoming calls are matched with the DNIS groups/call types assigned to customer profiles. If a match is found, the customer profile session and overflow limits are applied and if available, the required resources allocated. If a matching DNIS group/call type assigned to a customer profile is not found, the Default DNIS group/call type and default customer are used. Otherwise, the call is rejected.
After the call answered, Cisco RPMS checks the customer profile for an assigned VPDN group. If a VPDN group is found, Cisco RPMS matches the group's domain name or DNIS number with the incoming call. If a match is found, VPDN group session and overflow limits are applied and tunnel negotiation begins. If a match is not found or no VPDN group is assigned to the customer profile, the call is processed as retail dial service call if local AAA services are available or the call is rejected.
Figure 1-3 illustrates Cisco RPMS call processing.
This section describes the major features of Cisco RPMS:
A DNIS group is a configured list of DNIS numbers that correspond to the numbers dialed by particular customers, service offerings, or both. Cisco RPMS checks the DNIS number of inbound calls against the configured DNIS groups or the Default DNIS group. If a match is found, the configured information in the customer profile to which the DNIS group is assigned is used. If a match is not found, the Default DNIS group and default customer profile are used. If a customer profile is not configured, the call is rejected.
In the case of an 800, 888, or similar number, the DNIS number delivered to the NAS by the telco switch is different from the number dialed by the user. Cisco RPMS provides an optional reference number in each DNIS group for numbers that require public network database lookup and results in a POTS-routable DNIS.
Cisco RPMS enables you to maximize the use of available shared resources in a Cisco NAS or across multiple Cisco NASes for various resource allocation schemes. Cisco RPMS allows you to combine your Cisco NAS resource groups with call types (speech, digital, V.110, and V.120) and resource services for modems into resource data assignments. Resource data can be configured for customer profiles and assigned to incoming calls through DNIS groups and call types. You can also use the default DNIS group and a default customer profile to assign all incoming calls that do not match defined DNIS groups to resource data differentiated by call type.
Resource groups are used to partition the available resources in a NAS. For non-port based resources, the individual resource is identified by a port or a physical location, such as slot and port. Resource groups containing this type of resource explicitly list ranges of ports. Examples of this type of resource are modems or V.110 terminal adapters (TA).
For non-port based resources, resource groups have a single parameter identifying the size of the group. Examples of this type of resource are HDLC controllers for data calls or V.120 IOS-based TAs.
Resource groups are configured in the NAS. As a result, Cisco RPMS is configured to reference the NAS-configured resource groups and the Cisco RPMS resource group names must exactly match the NAS-configured resource groups.
| Command | Argument/Options |
|---|---|
Min-speed | <300-56000>, any |
Max-speed | <300-56000>, any |
Modulation | k56flex, v22bis,v32bis, v34, v90, any |
Error-correction | lapm, mnp4 |
Compression | mnp5, v42bis |
Cisco RPMS allows you to set session count and session overflow limits in each customer profile. The session count limit determines the maximum number of non-overflow sessions supported for a customer profile. When the session count limit is reached, if overflow sessions are not enabled, any new calls are rejected. If overflow sessions are enabled, new sessions up to the session overflow limit are processed and marked as overflow for call handling and accounting.
The session overflow limit determines the number of sessions above the session count limit that are allowed. If the session overflow limit is greater than zero, overflow sessions are enabled and the maximum number of allowed sessions is the session count limit plus the session overflow limit. When the session overflow limit is reached, any new calls are rejected.
Enabling overflow sessions can be useful for isolating excess sessions for preferred customers or premium rates. Overflow session can also be useful for encouraging customers to adequately forecast bandwidth usage or for special events when normal session usage is exceeded. For example, if a customer is having a corporate-wide program and many people are expected to request remote access, you could enable a large number of overflow sessions and charge a premium rate for the excess bandwidth requirements.
This table shows the effects of session count limit and session overflow limit settings combinations.
| Session Count Limit | Session Overflow Limit | Call Handling |
|---|---|---|
0 | 0 | Reject all calls. |
10 | 0 | Accept up to 10 sessions. |
10 | 10 | Accept up to 20 sessions. Mark sessions 11 - 20 as overflow sessions. |
0 | 10 | Accept up to 10 sessions and mark 1-10 session as overflow. |
All | 0 | Accept all calls. |
0 | All | Accept all calls and mark all calls as overflow. |
Cisco RPMS also allows you to set threshold settings and generate Email notification when thresholds for session count limit, session overflow limit, call rejections, and overflow call rejections are exceeded.
Cisco RPMS allows you to configure session count and session overflow limits per VPDN group so you can manage the VPDN sessions.
The session count limit determines the maximum number of non-overflow sessions supported for a VPDN group. When the session count limit is reached, if overflow sessions are not enabled, any new VPDN calls using the VPDN group sessions are rejected. If overflow sessions are enabled, new sessions up to the session overflow limit are processed and marked as overflow for call handling and accounting.
The VPDN session overflow limit determines the number of sessions above the session count limit that are allowed for the VPDN group. If the session overflow limit is greater than zero, overflow sessions are enabled and the maximum number of allowed sessions is the session count limit plus the session overflow limit. When the session overflow limit is reached, any new calls are rejected.
Enabling VPDN overflow sessions can be useful for isolating excess sessions for preferred customers or premium rates. Overflow sessions can also be useful for encouraging customers to adequately forecast bandwidth usage or for special events when normal session usage is exceeded. For example, if a customer is having a corporate-wide program and many people are expected to request remote access, you could enable a large number of overflow sessions and charge a premium rate for the excess bandwidth requirements.
To ensure home gateway resources are not consumed by a few users with MLP connections, Cisco RPMS also allows you to specify the maximum number of connections that can open multilink connections in a VPDN group. In addition, you can also specify the maximum number of links for each bundle. For example, if standard ISDN users will access the VPDN group, limit this setting to two links per bundle. If video conferencing will be used, increase this setting to accommodate the necessary bandwidth (usually six links).
Call treatment determines how calls are handled when certain events occur. For example, if the session and overflow limits for one of your customers has been exceeded, any additional calls will receive a busy signal.
Table 1-1 shows the types of events that require call treatment and the call treatment options.
| Event | Call Treatment |
|---|---|
Customer profile not found |
The call is rejected based on not matching a DNIS group/call type and customer profile. No answer, the default, implies that the NAS was appropriate, but resources were unavailable so the caller should try back later. Busy can be used to immediately reject the call and free up the circuit. |
Customer profile limits exceeded |
|
NAS resource not available |
The call can be answered, but the NAS does not have any available resources in the resource groups. Channel not available allows the switch to try additional channels until it gets to a different NAS in the same trunk group that has the available resources. Busy can be used when the trunk group does not span additional NASes. |
Call discrimination match |
|
Cisco RPMS can be configured to generate Email notification when the following configurable thresholds settings are exceeded:
From Customer Profile:
From VPDN Group:
From VPDN IP Endpoint:
After a threshold is exceeded, no additional Emails for that threshold are sent until the threshold is reactivated. The Alert LowWater Percentage parameter (Administration>Sever) specifies the percentage low point of the threshold that the value must return to in order to reactivate the threshold setting. This prevents Email storms when a threshold is continually reached without much variation.
Accounting data for network dial service usage can be generated in two ways:
For each call, information is sent from Cisco RPMS to the Cisco NAS specifying the resources used for the call.
The Cisco NAS can be configured to generate AAA accounting records and access an external AAA server option. The accounting start and stop records in AAA attribute format are sent to the external AAA server using either RADIUS or TACACS+ protocols for accounting data storage.
| Accounting Start Record | Accounting Stop Record |
|---|---|
Call-Type | Disconnect-Cause |
The call detail records contain additional information that is geared toward troubleshooting and capacity planning. For example, Cisco NASes do not generate an accounting record for calls that are not answered. In a resource pooling environment, detailed information regarding unanswered calls is important and can be used to signify that a customer's limit is too low (call not answered because customer limit is exceeded) or a Cisco NAS is low on modems.
Call detail record filenames use the following format:
cp-name_year-month-day_time.CSV
Call detail records store the following information for each call:
Cisco RPMS allows you to build fault tolerance and resiliency into your dial service offerings. Fault tolerance and resiliency consists of the following:
Cisco RPMS can run standalone or with a backup Cisco RPMS server. Each Cisco RPMS backup server can support multiple primary Cisco RPMS servers. The Cisco RPMS backup server requires Oracle replication and provides a backup configuration and counters in case the primary Cisco RPMS server(s) becomes unavailable. Identical configuration settings (customer profiles, DNIS groups, resource groups, VPDN groups, and others) must be configured on the primary and backup Cisco RPMS servers.
As calls are received, the primary Cisco RPMS server locally checks its session counts to perform session management. Periodically, these local counts are sent to the backup Cisco RPMS server for synchronization. When the session counts get close to a session limit, the primary Cisco RPMS server changes to get the session count from the backup server for each call to ensure an accurate session count is maintained in the primary and backup servers. When the session counts return to a lower level, Cisco RPMS goes back to local session counts to perform session management.
Database replication is accomplished using the Oracle Database Replication Manager and the Cisco RPMS DBServer component. All configuration is replicated among primary and backup Cisco RPMS servers. Configuration changes are automatically updated to peer Cisco RPMS servers through replication and the DBServer cache update mechanism.
Oracle replication ensures all Cisco RPMS databases are synchronized. However, because database replication is asynchronous, it does not maintain session counts. To ensure the backup Cisco RPMS knows the current session counts, the Distributed Session Manager (DSM) authority is maintained on the backup server.
Backup customer profiles are customer profiles configured locally on the Cisco NAS and used to answer calls based on a configured allocation scheme when the link between the Cisco NAS and Cisco RPMS becomes unavailable.
If identical configurations are used across multiple primary Cisco RPMS servers and the servers use the same backup Cisco RPMS server, Distributed Session Management (DSM) technology can be used to aggregate the resource and session counts on the backup Cisco RPMS server.
Cisco RPMS allows multiple administrators to manage resources and dial services. Administrators are identified by a username and password combination. Cisco RPMS allows you to add, edit, and delete administrators and administration rights with the Administration>Administrators command.
The following privileges are available for Cisco RPMS administrators. Each level includes the privileges of that level as well as all lower levels:
Call discrimination in Cisco RPMS allows you to specify DNIS groups and call types to block and disconnect calls before they are assigned to Cisco NAS resources or before any other Cisco RPMS processing occurs. For example, call discrimination can be used to restrict a specific DNIS group to only modem calls by creating call discrimination settings for the DNIS group and the other call types (digital, V.110, and V.120).
To configure call discrimination, select Configuration > Call Info > Discrimination. The Call Discrimination Table page appears. Enter a name for the call discrimination table entry in the CD Name field and select the DNIS groups and call types that will be prevented from accessing the Cisco NAS resources managed by this Cisco RPMS.
Repeat these steps for each DNIS group/call type combination that will be blocked.
Data over Voice Bearer Services (DOVBS) is a dial service that uses a customer profile and an assigned resource group of digital resources to direct data calls with a speech call type to HDLC controlers.
To support ISDN Data over Voice Bearer Service (DOVBS), use a DNIS group and a configured customer profile to direct the speech call to the appropriate digital resource.
The DNIS group that is assigned to the customer profile should have a call type of speech. The resource group assigned to this customer profile will be digital resources and also have a call type of speech, so the call will terminate on an HDLC controller rather than a modem.
Changes in Cisco RPMS are required when the Cisco NAS is connected to the PSTN by SS7 signaling and trunking.
Cisco RPMS provides pre-defined, system-wide resource groups to support SS7 calls. The desired SS7 resource group is configured for a customer's DNIS group/call type association. This DNIS group/call type/SS7 resource group association is then assigned to the customer profile. If an optional resource service is desired, also configure the SS7 resource group as part of the resource data assigned to the customer profile.
When assigning the desired DNIS Groups and call types to the customer profile, also assign the pre-defined SS7 resource group from the SS7 Resource Group pull-down menu using Configuration>Customer>DNIS Grps/Call Types&SS7. The SS7 Resource Group menu options are:
Use the None option to indicate there are no SS7 calls for the current DNIS group.
When SS7 and RPMS are enabled on the NAS:
One TACACS connection is required to enable the resource management protocol (RMP) for the NAS to communicate with Cisco RPMS. For example,
tacacs-server host 10.6.8.25 tacacs-server host 10.3.15.2 tacacs-server key cisco tacacs-server administration resource-pool aaa protocol tacacs
Remove any previously configured resource groups. For example,
no resource-pool group resource mica-res-group no resource-pool group resource isdn-res-group
Ensure resource pooling is still enabled.
To use RLM, you have to have a controller run PRI. For example, use the following commands:
controller T1 0 framing esf clock source line primary linecode b8zs pri-group timeslots 1-24
To start the RLM, perform the following configurations:
rlm group 3
isdn rlm 3
When using SS7 signaling and when the NAS is configured to communicate with Cisco RPMS, like resources in the NAS are put into single resource groups.
The following pre-defined resource groups are created on the NAS:
When these resource groups are configured, the NAS will not allow any customer configuration for new resource groups.
When the NAS receives the Call-Accept reply with the resource group name from Cisco RPMS, the NAS will try to allocate a resource from one of the pre-configured SS7 resource groups based on the following mapping:
| RPMS Resource Group | NAS Resource Group |
|---|---|
rg_ss7_rpms_speech | rg_ss7_v24_alog or rg_ss7_mica |
rg_ss7_rpms_digital | rg_ss7_digital |
rg_ss7_rpms_v110 | rg_ss7_mica or rg_ss7_v110 |
rg_ss7_rpms_v120 | rg_ss7_v120 |
Note that if the NAS needs to allocate a speech resource for an incoming call, the NAS will try rg_ss7_v24_analog first. Only when no resource is available from rg_ss7_v24_analog, will the NAS then try rg_ss7_mica. This same process is applied to the rg_ss7_v110 and rg_ss7_mica resource groups. For V.110 calls, the NAS looks for Toucan boards first and then Mica h/w with V.110 supported in p/w.
Once one matching resource is allocated, the resource-group name, from which the resource is allocated, is copied to Resource-Allocated message and sent to Cisco RPMS. This way Cisco RPMS knows which resource-group is used for the call.
When the Redundant Link Manager (RLM) is stopped, or Resource Management Protocl (RMP) is disabled on the NAS, the NAS puts the pre-configured SS7 resource groups back into system default resource group(s), removing them from use.
Any resource groups used for non-SS7 call processing must then be reconfigured on the NAS.
Cisco RPMS SS7 resource group configurations do not have to be changed. The Cisco RPMS SS7 resource group configurations can be used at the same time as the non-SS7 resource groups.
![]()
![]()
![]()
![]()
![]()
![]()
![]()
Posted: Fri Jul 2 13:08:59 PDT 1999
Copyright 1989-1999©Cisco Systems Inc.