Table Of Contents
gprs gtp echo-timer dynamic enable
gprs gtp echo-timer dynamic minimum
gprs gtp echo-timer dynamic smooth-factor
gprs gtp error-indication-throttle
gprs gtp ip udp ignore checksum
gprs gtp pdp-context timeout idle
gprs gtp pdp-context timeout session
gprs gtp ppp-regeneration vtemplate
gprs gtp response-message pco ipcp nack
gprs gtp response-message wait-accounting
gprs idle-pdp-context purge-timer
gprs maximum-pdp-context-allowed
gprs qos default-response requested
gprs radius attribute chap-challenge
gprs radius attribute session-timeout
gprs redundancy charging sync-window cdr rec-seqnum
gprs redundancy charging sync-window gtpp seqnum
gprs umts-qos map diffserv-phb
gprs umts-qos map traffic-class
gtp pdp-context single pdp-session
gtp pdp-context timeout session
gtp response-message wait-accounting
radius attribute acct-session-id charging-id
radius attribute suppress imsi
radius attribute suppress sgsn-address
radius attribute user-name msisdn
gprs gtp echo-timer dynamic enable
To enable the dynamic echo timer on the GGSN, use the gprs gtp echo-timer dynamic enable global configuration command. To disable the dynamic echo timer, use the no form of this command.
gprs gtp echo-timer dynamic enable
no gprs gtp echo-timer dynamic enable
Syntax Description
This command has no arguments or keywords.
Defaults
Disabled
Command Modes
Global configuration
Command History
Usage Guidelines
For a GTP path to be active, the SGSN needs to be active. To determine that an SGSN is active, the GGSN and SGSN exchange echo messages. Although the GGSN supports different methods of echo message timing, the basic echo flow begins when the GGSN sends an echo request message to the SGSN. The SGSN sends a corresponding echo response message back to the GGSN.
If the GGSN does not receive a response after a certain number of retries (a configurable value), the GGSN assumes that the SGSN is not active. This indicates a GTP path failure, and the GGSN clears all PDP context requests associated with that path.
The GGSN supports two different methods of echo timing—the default echo timer and the dynamic echo timer.
The GGSN's default echo timer can not be configured to accommodate network congestion and therefore the GTP path could be cleared prematurely. The dynamic echo timer feature enables the GGSN to better manage the GTP path during periods of network congestion. Use the gprs gtp echo-timer dynamic enable command to enable the GGSN to perform dynamic echo timing.
Default echo timer
The dynamic echo timer is based on the default echo timer in the GGSN. A description of the default echo timer follows as a means of comparison.
The default echo timer configuration uses the following commands:
•
gprs gtp n3-requests—Specifies maximum number of times that the GGSN attempts to send a echo-request message. The default is 5 times.
•
•
If the GGSN receives the echo response within the path echo interval (as specified in the gprs gtp path-echo-interval command; default is 60 seconds), it sends another echo request message after 60 seconds (or whatever time was configured in the gprs gtp path-echo-interval command). This message flow continues as long as the GGSN receives an echo response message within the specified path echo interval.
If the GGSN fails to receive an echo response message within the path echo interval, it resends echo request messages until the N3-requests counter is reached (as specified by the gprs gtp n3-requests command; default is 5). Because the initial request message is included in the N3-requests counter, the total number of retries is N3-1. The T3 timer increases by a factor of two for each retry (the factor value is not configurable).
For example, if N3 is set to the default of 5, and T3 is set to the default of 1 second, the GGSN will resend 4 echo request messages (the initial request + 4 retries=5). The T3 time increments for each additional echo request, by a factor of 2 seconds. So, the GGSN resends a message in 2 seconds, 4 seconds, 8 seconds, and 16 seconds. If the GGSN fails to receive an echo response message within the time period of the N3-requests counter, it clears the GTP path and deletes all of the PDP contexts.
For the above example, the total elapsed time from when the first request message is sent, to when the GTP path is cleared, is: 60+2+4+8+16=90 seconds,
where 60 is the initial value of the path echo interval, and the remaining 4 time periods are the increments of the T3 timer for the subsequent retries.
Dynamic echo timer
The dynamic echo timer method is different from the default echo timer method on the GGSN because it uses a calculated round-trip timer (RTT), as well as a configurable factor or multiplier to be applied to the RTT statistic.
The dynamic echo timer configuration uses the following commands:
•
•
•
•
•
The GGSN calculates the RTT statistic for use by the dynamic echo timer feature. The RTT is the amount of time between sending a particular echo request message and receiving the corresponding echo response message. RTT is calculated for the first echo response received; the GGSN records this statistic. Because the RTT value might be a very small number, there is a minimum time for the dynamic echo timer to use. This value is configured using the gprs gtp echo-timer dynamic minimum command.
If the GGSN fails to receive an echo response message within the path echo interval, it goes into retransmission, or path failure mode. During path failure mode, the GGSN uses a value referred to as the T-dynamic. The T-dynamic is the greater of either the dynamic minimum, or the RTT statistic multiplied by the smooth factor.
The T-dynamic essentially replaces the use of the gprs gtp t3-response command, which is used in the default echo timer method on the GGSN. The T-dynamic timer increases by a factor of two for each retry (again, this factor is not configurable), until the N3-requests counter is reached (N3-requests counter includes the initial request message).
For example, if the RTT is 6 seconds, N3 is set to 5, and the smooth factor is set to 3, the GGSN will resend 4 echo request messages in path failure mode. The T-dynamic value is 18 (RTT x smooth factor), so the GGSN sends a retry echo request message in 36 seconds, 72 seconds, 144 seconds, and 288 seconds. If the GGSN fails to receive an echo response message in this time period, it clears the GTP path and deletes all PDP contexts. The total elapsed time from when the first request message is sent to when the GTP path is cleared is: 60+36+72+144+288=600 seconds,
where 60 is the initial value of the path echo interval, and the remaining 4 time periods are the increments of the T-dynamic for the subsequent retries.
Examples
The following example turns on the dynamic echo timer, sets the minimum value to 5 seconds, and configures a smooth factor of 3:
gprs gtp echo-timer dynamic enablegprs gtp echo-timer dynamic minimum 5gprs gtp echo-timer dynamic smooth-factor 3Related Commands
gprs gtp echo-timer dynamic minimum
To specify the minimum time period used by the dynamic echo timer, use the gprs gtp echo-timer dynamic minimum global configuration command. To return to the default value, use the no form of this command.
gprs gtp echo-timer dynamic minimum number
no gprs gtp echo-timer dynamic minimum number
Syntax Description
number
Minimum time period (between 1 and 60 seconds) of the dynamic echo timer. Value must be an integer. The default value is 5 seconds.
Defaults
5 seconds
Command Modes
Global configuration
Command History
Usage Guidelines
Use this command to specify the minimum time period (in seconds) used by the dynamic echo timer, also referred to as the T-dynamic. If the GGSN's current calculation of the round-trip timer (RTT) statistic, multiplied by the smooth factor, is less than the configured dynamic minimum value, then the GGSN uses the configured minimum as the T-dynamic.
The GGSN calculates the RTT statistic for use by the dynamic echo timer feature. The RTT is the amount of time between sending a particular echo request message and receiving the corresponding echo response message. RTT is calculated for the first echo response received; the GGSN records this statistic. Because the RTT value might be a very small number, there is a minimum time for the dynamic echo timer to use. This value is configured using the gprs gtp echo-timer dynamic minimum command.
If the GGSN fails to receive an echo response message from the SGSN within the path echo interval, it goes into retransmission, or path failure mode. During path failure mode, the GGSN uses a value referred to as the T-dynamic. The T-dynamic is the greater of either the dynamic minimum, or the RTT statistic multiplied by the smooth factor.
The T-dynamic essentially replaces the use of the gprs gtp t3-response command, which is used in the default echo timer method on the GGSN. The T-dynamic timer increases by a factor of two for each retry (again, this factor is not configurable), until the N3-requests counter is reached (N3-requests counter includes the initial request message).
Note
Examples
The following example turns on the dynamic echo timer, sets the minimum value to 6 seconds, and configures a smooth factor of 2:
gprs gtp echo-timer dynamic enablegprs gtp echo-timer dynamic minimum 6gprs gtp echo-timer dynamic smooth-factor 2Related Commands
gprs gtp echo-timer dynamic smooth-factor
To configure the multiplier that the GGSN uses to calculate the time to wait to send retries of the dynamic echo timer, use the gprs gtp echo-timer dynamic smooth-factor global configuration command. To return to the default value, use the no form of this command.
gprs gtp echo-timer dynamic smooth-factor number
no gprs gtp echo-timer dynamic smooth-factor number
Syntax Description
number
Integer (between 1 and 100) used by the GGSN as a multiplier for the RTT statistic, to calculate the T-dynamic. The default is 2.
Defaults
2
Command Modes
Global configuration
Command History
Usage Guidelines
The dynamic echo timer uses the smooth factor to calculate what is known as the T-dynamic. The T-dynamic is calculated by multiplying the RTT (or the value configured in the gprs gtp echo-timer dynamic minimum, whichever is greater) times the smooth-factor.
Note
Examples
The following example turns on the dynamic echo timer, sets the minimum value to 1 second, and configures a smooth factor of 2:
gprs gtp echo-timer dynamic enablegprs gtp echo-timer dynamic minimum 1gprs gtp echo-timer dynamic smooth-factor 2Related Commands
gprs gtp error-indication-throttle
To specify the maximum number of error indication messages that the GGSN sends out in one second, use the gprs gtp error-indication-throttle command. To return to the default value, issue the no form of this command.
gprs gtp error-indication-throttle window-size size
no gprs gtp error-indication-throttle
Syntax Description
size
Integer (between 0 and 256) that specifies the maximum number of error indication messages that the GGSN sends in one second.
Defaults
Error indication throttling is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
GTP error indication messages are sent by the GGSN to the SGSN when the SGSN sends data for PDP context the GGSN cannot locate. The error indication message informs the SGSN that the PDP context cannot be located so that the SGSN can clean up the PDP context on its end.
Use the gprs gtp error-indication-throttle command to specify the maximum number of error indication messages that are sent by the GGSN in one second. This provides a way to implement flow control for transmission of GTP error messages. This command sets the initial value of a counter which is decremented each time an error indication message is sent. When the counter reaches zero, the GGSN stops transmitting error indication messages. The GGSN resets this counter to the configured throttle value after one second.
If you do not issue the command, error indication throttling is not enabled. To restore the default value (error indication throttling is disabled) use the no form of this command.
Examples
The following example shows a throttle value of 150:
gprs gtp error-indication-throttle window-size 150gprs gtp ip udp ignore checksum
To configure the GGSN to ignore user datagram protocol (UDP) checksums (in order to support CEF switching on the GGSN), use the gprs gtp ip udp ignore checksum global configuration command. To disable the ignoring of UDP checksums on the GGSN, use the no form of this command.
gprs gtp ip udp ignore checksum
no gprs gtp ip udp ignore checksum
Syntax Description
This command has no arguments or keywords.
Defaults
Releases prior to Cisco IOS Release 12.3(14)XU, UDP checksums are verified by default.
With Cisco IOS Release 12.3(14)XU and later, UDP checksums are ignored by default.
Command Modes
Global configuration
Command History
Usage Guidelines
UDP checksum verification can prohibit operation of CEF switching processing on the GGSN if the checksum should have a non-zero result. Therefore, if you want to enable CEF switching on the GGSN, ensure that the GGSN is configured to ignore UPD checksums (the default).
If UDP checksum verification remains enabled on the GGSN and a non-zero result occurs, the GTP T-PDUs will be process switched, even if you have configured the GGSN for CEF switching.
The gprs gtp ip udp ignore checksum command does not apply if you are only using process switching on the GGSN.
Note
For more information about switching processes, refer to the Cisco IOS Switching Services Configuration Guide.
Examples
The following example disables UDP checksum verification on the GGSN:
gprs gtp ip udp ignore checksumRelated Commands
gprs gtp map signalling tos
To specify an IP ToS mapping for GPRS tunneling protocol (GTP) signaling packets, use the gprs gtp map signalling tos global configuration command. To return to the default value, use the no form of this command.
gprs gtp map signalling tos tos-value
no gprs gtp map signalling tos tos-value
Syntax Description
Defaults
ToS value 5
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs gtp map signalling tos command to specify the IP ToS mapping for GTP signaling packets transmitted by the GGSN. The higher the value, the higher the class of service provided to the packets.
Examples
The following example specifies a IP ToS mapping value of 3:
gprs gtp map signalling tos 3Related Commands
gprs gtp n3-buffer-size
To specify the size of the receive buffer that the GGSN uses to receive GTP signaling messages and packets sent through the tunneling protocol, use the gprs gtp n3-buffer-size global configuration command. To return to the default value, use the no form of this command.
gprs gtp n3-buffer-size bytes
no gprs gtp n3-buffer-size
Syntax Description
bytes
Number of bytes (between 2048 and 65535) that specifies the size of the N3 buffer. The default is 8192 bytes.
Defaults
8192 bytes
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs gtp n3-buffer-size command to specify the size of the GTP N3 buffer on the GGSN. The N3 buffer is a receive buffer that the GGSN uses to receive GTP signaling messages and packets sent through the tunneling protocol. The recommended value for the N3 buffer size is 8192 (the default size).
Examples
The following example specifies a buffer size of 2084 bytes:
gprs gtp n3-buffer-size 2048gprs gtp n3-requests
To specify the maximum number of times that the GGSN attempts to send a signaling request to an SGSN, use the gprs gtp n3-requests global configuration command. To return to the default value, use the no form of this command.
gprs gtp n3-requests requests
no gprs gtp n3-requests requests
Syntax Description
requests
A number between 1 and 65535 that specifies the number of times a request is attempted. The default is 5 requests.
Defaults
5 requests
Command Modes
Global configuration
Command History
Usage Guidelines
The value of the gprs gtp n3-requests command is used for all signaling requests on the GGSN.
The GGSN supports two different methods of echo timing—the default echo timer and the dynamic echo timer. The gprs gtp n3-requests command is used by the GGSN to perform either type of echo processing.
Examples
The following example shows the GGSN attempting to send a signaling request 3 times:gprs gtp n3-requests 3Related Commands
gprs gtp path-echo-interval
To specify the number of seconds that the GGSN waits before sending an echo-request message to the SGSN or charging gateway, use the gprs gtp path-echo-interval global configuration command. To return to the default value, use the no form of this command.
gprs gtp path-echo-interval interval
no gprs gtp path-echo-interval interval
Syntax Description
Defaults
60 seconds
Command Modes
Global configuration
Command History
Usage Guidelines
The GGSN supports two different methods of echo timing—the default echo timer and the dynamic echo timer. The gprs gtp path-echo-interval command is used on the GGSN to perform either type of echo processing.
Use the gprs gtp path-echo-interval command to specify the interval that the GGSN waits before sending an echo-request message to the SGSN or charging gateway to check for GTP path failure.
Note
Examples
The following example shows the GGSN waiting 90 seconds before sending an echo-request message:
gprs gtp path echo-interval 90Related Commands
gprs gtp pdp-context timeout idle
To specify the time, in seconds, that a GGSN allows a session to remain idle at any access point before purging the PDP context, use the gprs gtp pdp-context timeout idle global configuration command. To return to the default value, use the no form of this command.
gprs gtp pdp-context timeout idle seconds [uplink]
no gprs gtp pdp-context timeout idle
Syntax Description
Defaults
259200 seconds (72 hours)
Command Modes
Global configuration
Command History
Usage Guidelines
The GGSN supports the RADIUS Idle-Timeout (Attribute 28) field. The GGSN stores the attribute 28 value if it is present in the access request packets sent by the AAA server. When a PDP context is idle for an amount of time that exceeds the session idle timeout duration, the GGSN terminates it.
The duration specified for the session idle timer applies to all PDP contexts of a session, however, a session idle timer is started for each PDP context. Therefore, the session idle timer is per-PDP, but the timer duration is per-session.
On the GGSN, the session idle timer can be configured globally and at the APN. The value configured at the APN level using the gtp pdp-context timeout idle access-point configuration command overrides the value configured globally using the gprs gtp pdp-context timeout idle global configuration command. The value configured in the user profile on the RADIUS server overrides the value configured at the APN.
Note
You can disable the session idle timer for a particular user by configuring 0 as the session idle time duration in the user profile on the RADIUS server. If a user is authenticated by RADIUS, the session idle time cannot be disabled.
Note
Note
Examples
The following example shows configuring the GGSN to wait 18000 seconds before ending an idle PDP context:
gprs gtp pdp-context timeout idle 18000Related Commands
gprs gtp pdp-context timeout session
To specify the time, in seconds, that the GGSN allows a session to exist at any access point before terminating the session, use the gprs gtp pdp-context timeout session global configuration command. To return to the default value, use the no form of this command.
gprs gtp pdp-context timeout session seconds
no gprs gtp pdp-context timeout session
Syntax Description
seconds
Time, in seconds, that the GGSN allows a session to exist at any access point. Specify a value between 30 and 4294967 seconds.
Defaults
Disabled
Command Modes
Global configuration
Command History
Usage Guidelines
When enabled using the gprs radius attribute session-timeout command, the GGSN supports the RADIUS Session-Timeout (Attribute 27). The GGSN stores the attribute timeout value received in access-accept packets sent by the AAA server and when the duration of a session exceeds the duration configured as absolute session timer, the GGSN terminates the session and all PDP contexts belonging to the session (those with the same IMSI or MS address).
Note
Note
On the GGSN, the absolute session timer can be configured globally and at the APN. The value configured at the APN level using the gtp pdp-context timeout session access-point configuration command overrides the value configured globally using the gprs gtp pdp-context timeout session global configuration command. The value configured in the user profile on the RADIUS server overrides the value configured at the APN.
Examples
The following example shows configuring the GGSN to end any session that exceeds 86400 seconds in duration:
gprs gtp pdp-context timeout session 86400Related Commands
gprs gtp ppp vtemplate
To associate the virtual template interface that defines the PPP characteristics with support for the PPP PDP type over GTP on the GGSN, use the gprs gtp ppp vtemplate global configuration command. To remove specification of the PPP virtual template interface for GTP on the GGSN, use the no form of this command.
gprs gtp ppp vtemplate number
no gprs gtp ppp vtemplate
Syntax Description
Defaults
No default behavior or values.
Command Modes
Global configuration
Command History
Usage Guidelines
Before you configure the gprs gtp ppp vtemplate command, you must configure the virtual template interface with the necessary PPP characteristics. The number that you configure for the virtual template interface that defines the PPP characteristics, must correspond to the number that you specify in the gprs gtp ppp vtemplate command.
Examples
The following example configures two virtual template interfaces on the GGSN, one for GTP encapsulation and one for PPP, and specifies the PPP virtual template interface for GTP on the GGSN.
Note
The first section of commands configures the GPRS virtual template interface for GTP:
interface Virtual-Template 1ip unnumber loopback 1no ip directed-broadcastencapsulation gtpno ip route-cachegprs access-point-list gprsThe following example configures a virtual template interface for PPP and associates the virtual template for support of the PPP PDP type over GTP on the GGSN:
interface Virtual-Template 2ip unnumbered FastEthernet 1/0no ip directed-broadcastno peer default ip addressppp authentication chapppp timeout retry 30gprs gtp ppp vtemplate 2Related Commands
interface virtual-template
Creates a virtual template interface that can be configured and applied dynamically in creating virtual access interfaces.
gprs gtp ppp-regeneration vtemplate
To associate the virtual template interface that is configured for PPP encapsulation with support for regenerated PPP sessions on the GGSN, use the gprs gtp ppp-regeneration vtemplate global configuration command. To remove specification of the PPP virtual template interface for regenerated PPP sessions on the GGSN, use the no form of this command.
gprs gtp ppp-regeneration vtemplate number
no gprs gtp ppp-regeneration vtemplate
Syntax Description
Defaults
No default behavior or values.
Command Modes
Global configuration
Command History
Usage Guidelines
Before you configure the gprs gtp ppp-regeneration vtemplate command, you must configure the virtual template interface for PPP encapsulation using the encapsulation ppp command. In addition, you must also configure the ip address negotiated command and the no peer neighbor-route command at the virtual template interface for PPP encapsulation.
The number that you configure for the virtual template interface to support PPP encapsulation, must correspond to the number that you specify in the gprs gtp ppp-regeneration vtemplate command.
Examples
The following example configures two virtual template interfaces on the GGSN, one for GTP encapsulation for communication between the GGSN and the SGSN, and one for PPP regeneration. The virtual template interface for PPP regeneration supports the creation of PPP sessions from the GGSN over Layer 2 Tunneling Protocol (L2TP) tunnels to an L2TP network server (LNS).
Note
The first section of commands configures the GPRS virtual template interface for GTP:
interface Virtual-Template 1ip unnumber loopback 1no ip directed-broadcastencapsulation gtpno ip route-cachegprs access-point-list gprsThe following example configures a virtual template interface for PPP regeneration:
interface Virtual-Template 11ip address negotiatedno peer neighbor-routeencapsulation ppp
Note
The following example specifies virtual template interface 11 for PPP regeneration on the GGSN:
gprs gtp ppp-regeneration vtemplate 11Related Commands
interface virtual-template
Creates a virtual template interface that can be configured and applied dynamically in creating virtual access interfaces.
gprs gtp response-message pco ipcp nack
To configure the GGSN to return an IPCP Conf-Nack (Code 03) in the GTP protocol configuration option (PCO) information element (IE) of a create PDP context response when returning IP Control Protocol (IPCP) options for which the granted values (non-zero) differ from those requested (IPCP Conf-Reject [Code 04] for those options for which the returned address values are zero), use the gprs gtp response-message pco ipcp nack global configuration command. To return to the default, use the no form of the command.
gprs gtp response-message pco ipcp nack
no gprs gtp response-message pco ipcp nack
Syntax Description
This command has no arguments or keywords.
Defaults
The GGSN sends an IPCP Conf-Ack (Code 2) in the PCO IE of the create PDP context response for the IPCP options for all the requested IPCP address options supported by the GGSN. The values being returned might be the same as or differ from those requested, or be zero.
For unsupported options, an IPCP Conf-Reject is returned.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs gtp response-message pco ipcp nack command to configure the GGSN to return an IPCP Conf-Nack in the PCO IE of a create PDP context response when returning IPCP options for which the granted values differ from those requested.
When the gprs gtp response-message pco ipcp nack command is configured, and the PCO IE of the create PDP context request contains IPCP options, the PCO IE in the create PDP response includes the following, depending on the whether options are supported by (and values are acceptable to) the GGSN:
•
•
•
gprs gtp response-message wait-accounting
To configure the GGSN to wait for a RADIUS accounting response before sending a create PDP context response to the SGSN for create PDP context requests received across all access points, use the gprs gtp response-message wait-accounting global configuration command. To configure the GGSN to send a create PDP context response to the SGSN after sending a RADIUS start accounting message to the RADIUS server (without waiting for a response from the RADIUS accounting server), use the no form of this command.
gprs gtp response-message wait-accounting
no gprs gtp response-message wait-accounting
Syntax Description
This command has no arguments or keywords.
Defaults
The GGSN sends a create PDP context response to the SGSN after sending a RADIUS start accounting message to the RADIUS accounting server. The GGSN does not wait for a RADIUS accounting response from the RADIUS accounting server.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs gtp response-message wait-accounting command to configure the GGSN to wait for a RADIUS accounting response from the RADIUS accounting server before sending a create PDP context response to the SGSN for create PDP context requests received across all access points.
If the GGSN does not receive a response from the RADIUS accounting server when you have configured the gprs gtp response-message wait-accounting command, it rejects the PDP context request.
When broadcast accounting is used (accounting requests are sent to multiple RADIUS servers), if a RADIUS server responds with an accounting response, the GGSN sends a create PDP context response and does not wait for the other RADIUS servers to respond.
The GGSN supports configuration of RADIUS response message waiting at both the global and access-point configuration levels. You can minimize your configuration by specifying the configuration that you want to support across most APNs, at the global configuration level. Then, at the access-point configuration level, you can selectively modify the behavior that you want to support at a particular APN. Therefore, at the APN configuration level, you can override the global configuration of RADIUS response message waiting.
To configure the GGSN to wait for a RADIUS accounting response as the default behavior for all APNs, use the gprs gtp response-message wait-accounting global configuration command. To disable this behavior for a particular APN, use the no response-message wait-accounting access-point configuration command.
To verify whether RADIUS response message waiting is enabled or disabled at an APN, you can use the show gprs access-point command and observe the value reported in the wait_accounting output field.
Examples
The following example globally configures the GGSN to wait for a RADIUS accounting response from the RADIUS accounting server before sending an activate PDP context response to the SGSN, for PDP context requests received across all access points except access-point 1. RADIUS response message waiting has been overridden at access-point 1 using the no gtp response-message wait-accounting command:
Note
aaa new-model!aaa group server radius fooserver 10.2.3.4server 10.6.7.8!aaa authentication ppp foo group fooaaa authorization network default group radiusaaa accounting exec default start-stop group foo!gprs access-point-list gprsaccess-point 1access-mode non-transparentaccess-point-name www.pdn1.comaaa-group authentication foono gtp response-message wait-accountingexitaccess-point 2access-mode non-transparentaccess-point-name www.pdn2.comaaa-group authentication foo!gprs gtp response-message wait-accounting!radius-server host 10.2.3.4 auth-port 1645 acct-port 1646 non-standardradius-server host 10.6.7.8 auth-port 1645 acct-port 1646 non-standardradius-server key ggsntelRelated Commands
gprs gtp t3-response
To specify the initial time that the GGSN waits before resending a signaling request message when a response to a request has not been received, use the gprs gtp t3-response global configuration command. To return to the default value, use the no form of this command.
gprs gtp t3-response response-interval
no gprs gtp t3-response response-interval
Syntax Description
response-interval
A value between 1 and 65535 that specifies the length of the T3 response interval, in seconds. The default is 1 second.
Defaults
1 second
Command Modes
Global configuration
Command History
Usage Guidelines
The gprs gtp t3-response command is used by the GGSN to process delete PDP context requests and to perform the default method of echo timing.
For delete PDP context requests, the gprs gtp t3-response command is used by the GGSN to specify how long the GGSN waits before sending a retry of the delete PDP context request when a response is not received from the SGSN, until the gprs gtp n3-requests limit is reached.
The GGSN supports two echo timer implementations—the default echo timer and the dynamic echo timer. The gprs gtp t3-response command also is used on the GGSN to perform the default type of echo processing, when the dynamic echo timer is not enabled.
If the GGSN receives the echo response within the path echo interval (as specified in the gprs gtp path-echo-interval command; default is 60 seconds), it sends another echo request message after 60 seconds (or whatever time was configured in the gprs gtp path-echo-interval command). This message flow continues as long as the GGSN receives an echo response message within the specified path echo interval.
If the GGSN fails to receive an echo response message from the SGSN within the path echo interval, it resends echo request messages until the N3-requests counter is reached (as specified by the gprs gtp n3-requests command; default is 5). Because the initial request message is included in the N3-requests counter, the total number of retries is N3-1. The T3 timer increases by a factor of two for each retry (the factor value is not configurable).
For example, if N3 is set to the default of 5, and T3 is set to the default of 1 second, the GGSN will resend 4 echo request messages (the initial request + 4 retries=5). The T3 time increments for each additional echo request, by a factor of 2 seconds. So, the GGSN resends a message in 2 seconds, 4 seconds, 8 seconds, and 16 seconds. If the GGSN fails to receive an echo response message from the SGSN within the time period of the N3-requests counter, it clears the GTP path and deletes all of the PDP contexts.
For the above example, the total elapsed time from when the first request message is sent, to when the GTP path is cleared, is: 60+2+4+8+16=90 seconds,
where 60 is the initial value of the path echo interval, and the remaining 4 time periods are the increments of the T3 timer for the subsequent retries.
Examples
The following example shows a T3 interval response interval of 524 seconds:
gprs gtp t3-response 524Related Commands
gprs idle-pdp-context purge-timer
To specify the time, in hours, that the GGSN waits before purging idle mobile sessions, use the gprs idle-pdp-context purge-timer global configuration command. To return to the default value, use the no form of this command.
gprs idle-pdp-context purge-timer hours
no gprs idle-pdp-context purge-timer hours
Syntax Description
hours
Value between 0 and 255 that specifies the number of hours that the GGSN waits before purging idle sessions. The value 0 disables the purge timer. The default is 72 hours.
Defaults
72 hours
Command Modes
Global configuration
Command History
Usage Guidelines
To specify the time that the GGSN waits before purging idle mobile sessions, use the gprs idle-pdp-context purge-timer command. To disable this feature, specify a purge-timer value of 0.
You can override the value of the global purge timer using the session idle-time access-point configuration command.
Note
Examples
The following example specifies that the GGSN wait for 60 hours before purging idle sessions:
gprs idle-pdp-context purge-timer 60Related Commands
gprs maximum-pdp-context-allowed
To specify the maximum number of PDP contexts (mobile sessions) that can be activated on the GGSN, use the gprs maximum-pdp-context-allowed global configuration command. To return to the default value, use the no form of this command.
gprs maximum-pdp-context-allowed pdp-contexts
no gprs maximum-pdp-context-allowed pdp-contexts
Syntax Description
pdp-contexts
Integer between 1 and 4294967295 that specifies the number of active PDP contexts allowed. The default is 10000 PDP contexts.
Defaults
10000 PDP contexts
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs maximum-pdp-context-allowed command to specify the maximum number of PDP contexts allowed on the GGSN. When the maximum allowable number of PDP contexts is reached, the GGSN refuses new PDP contexts (mobile sessions) until sessions are available.
The practical upper limit for the maximum number of PDP contexts supported on a GGSN is dependent on the memory and platform in use and the GGSN configuration (for example, whether or not a method of Point to Point Protocol [PPP] has been configured to forward packets beyond the terminal equipment and mobile termination, whether Dynamic Feedback Protocol [DFP] is being used or the memory protection feature is enabled, and the rate of PDP context creation to be supported).
Note
Cisco 7200 Series Router
The following list shows the maximum number of PDP contexts supported on the GGSN according to the memory and Cisco 7206 router series in use when a method of PPP has not been configured:
•
•
Catalyst 6500 Series Switch / Cisco 7600 Series Router
The Cisco MWAM can support up to 60,000 IP PDP contexts per GGSN instance with a maximum number of 300,000 IP PDP contexts per MWAM on which five GGSNs are configured.
Note
Note
DFP weighs PPP PDPs against IP PDPs with one PPP PDP being equal to 8 IP PDPs. therefore when using DFP, be aware that the configured maximum number of PDP contexts affects the GGSN weight. The lower the maximum number of PDP contexts, the lower the weight when all other parameters remain the same.
Note
Examples
In the following example 15000 PDP contexts are allowed on the GGSN:
gprs maximum-pdp-context-allowed 15000Related Commands
gprs idle-pdp-context purge-timer
Specifies the time that the GGSN waits before purging idle mobile sessions.
gprs mcc mnc
To configure the mobile country code and mobile network code that the GGSN uses to determine if a create PDP context request is from a roamer, use the gprs mcc mnc global configuration command. To return to the default values, use the no form of this command.
gprs mcc mcc-num mnc mnc-num [trusted]
no gprs mcc mcc-num mnc mnc-num [trusted]
Syntax Description
Defaults
000—For both the MCC and MNC. A valid code must be a non-zero value.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs mcc mnc command as part of the configuration required on the GGSN to support creation of CDRs for roaming mobile subscribers, or to block roamers from being able to create PDP context requests.
The MCC and MNC together identify a GPRS/UMTS PLMN. The values you configure using the gprs mcc mnc command without the trusted keyword option specified are those of the home PLMN ID - the PLMN to which the GGSN belongs. Only one home PLMN can be defined for a GGSN at a time. The GGSN uses the values that you configure in this command to compare with the IMSI in a create PDP context request.
The GGSN automatically specifies values of 000 for the MCC and MNC. However, you must configure non-zero values for both the MCC and MNC before you can enable the GGSN to create charging CDRs for roamers.
To properly issue the gprs mcc mnc command, you must specify both the mcc keyword with its argument and the mnc keyword with its argument. You cannot issue the command without specifying both keywords.
It is important that you configure the gprs mcc mnc and gprs charging roamers commands in their proper order. After you configure the MCC and MNC values, use the gprs charging roamers command to enable charging for roamers on the GGSN. You can change the MCC and MNC values by reissuing the gprs mcc mnc command.
Using the gprs mcc mnc command, you can also configure up to 5 "trusted" PLMNs by specifying the trusted keyword. A create PDP context request from a mobile subscriber in a trusted PLMN is treated the same as a create PDP context request from a mobile subscriber in the home PLMN.
To verify your configuration of these codes on the GGSN, use the show gprs charging parameters command.
Note
Examples
The following example replaces the default values of 000 on the GGSN, and specifies an MCC code of 310 for the USA and an MNC code of 15 for the Bell South service provider:
gprs mcc 310 mnc 15Related Commands
gprs memory threshold
To prevent the GGSN from draining processor memory during abnormal conditions (such as charging gateways [CGs] being down), use the gprs memory threshold global configuration command to configure a memory threshold, that when reached, activates the memory protection feature on the GGSN.
gprs memory threshold threshold
Syntax Description
threshold
Memory threshold, that when fallen below enables the memory protection feature on the GGSN. Valid range is 0 to 1024.
Defaults
The default is 10% of the total memory available at the time GGSN services are enabled.
Command Modes
Global configuration
Command History
Usage Guidelines
The GGSN memory protection feature prevents processor memory from being drained during periods of abnormal conditions (such as when all charging gateways are down and the GGSN is buffering CDRs into memory.
By default, the memory threshold is 10% of the total memory available at the time GGSN services are enabled using the gprs ggsn service global configuration command. You can use the gprs memory threshold global configuration command to configure the threshold according to the router and memory size.
When the amount of memory remaining on the system reaches the defined threshold, the memory protection feature activates and the GGSN performs the following actions to keep the processor memory from falling below the threshold:
•
•
•
Examples
The following example sets the memory threshold to 50 KB:
gprs memory threshold 512gprs ms-address exclude-range
To specify the IP address range(s) used by the GPRS network, and thereby excluded from the mobile station (MS) IP address range, use the gprs ms-address exclude-range global configuration command. To remove the specified range(s), use the no form of this command.
gprs ms-address exclude-range start-ip end-ip
no gprs ms-address exclude-range start-ip end-ip
Syntax Description
Defaults
No default behavior or values.
Command Modes
Global configuration
Command History
Usage Guidelines
An MS can not have the same IP address as another GPRS network entity. Use the gprs ms-address exclude-range command to reserve certain IP address ranges for use by the GPRS network, and to disallow them from use by an MS.
The gprs ms-address exclude range command verification is performed only for IP PDPs and does not apply to MS addresses assigned to VPNs or for PPP-Regen or PPP PDP types.
During a create PDP context request, the GGSN verifies whether the IP address of an MS falls within the specified excluded range. If there is an overlap of the MS IP address with an excluded range, then the create PDP context request is rejected. This measure prevents duplicate IP addressing in the network.
You can configure up to 100 IP address ranges. A range can be one or more addresses. However, you can configure only one IP address range per command entry. To exclude a single IP address, you can repeat the IP address in the start-ip and end-ip arguments. IP addresses are 32-bit values.
Examples
Example 1
The following example specifies the IP address ranges used by the GPRS network (which are thereby excluded from the MS IP address range:
gprs ms-address exclude-range 10.0.0.1 10.20.40.50gprs ms-address exclude-range 172.16.150.200 172.30.200.255gprs ms-address exclude-range 192.168.100.100 192.168.200.255Example 2
The following example excludes an MS from using the IP address of 10.10.10.1:
gprs ms-address exclude-range 10.10.10.1 10.10.10.1Related Commands
show gprs ms-address exclude-range
Displays the IP address range(s) configured on the GGSN for the GPRS network.
gprs plmn ip address
To specify the IP address range of a PLMN, use the gprs plmn ip address global configuration command.
gprs plmn ip address start_ip end_ip [sgsn]
Syntax Description
Defaults
No default behavior or values.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs plmn ip address global configuration command to specify the IP address range of the PLMN.
The gprs plmn ip address command defines addresses that belong to a PLMN. To indicate that the addresses are SGSN addresses within the PLMN, issue the gprs plmn ip address command with the sgsn keyword option specified. This option is used by the charging for roamers feature (gprs charging roamers command).
When using the gprs plmn ip address command with the GGSN charging for roamers feature, depending on how the PLMN IP address ranges have been defined using the gprs plmn ip address start_ip end_ip [sgsn] command, the charging for roamers feature operates as follows:
•
•
With this configuration, the following scenarios outline how the charging for roamers feature will function:
–
–
Configuration Guidelines
To enable charging for roamers on the GGSN, you should first define a set of IP address ranges for a PLMN using the gprs plmn ip address command.
It is important that you configure the gprs plmn ip address and gprs charging roamers commands in their proper order. After you configure the IP address range for a PLMN, use the gprs charging roamers command to enable charging for roamers on the GGSN. You can change the IP address range by reissuing the gprs plmn ip address command.
To verify your configuration, use the show gprs charging parameters command to see if the charging for roamers command is enabled. To verify your PLMN IP address ranges, use the show gprs plmn ip address command.
Examples
The following example specifies the IP address range of a PLMN:
gprs plmn ip address 10.0.0.1 10.20.40.50Related Commands
gprs charging roamers
Enables charging for roamers on the GGSN.
show gprs plmn ip address
Displays a list of IP address ranges defined for the PLMN.
gprs qos bandwidth-pool
To create or modify a Call Admission Control (CAC) bandwidth pool that can be attached to one or more APNs, use the gprs qos bandwidth-pool global configuration command. To delete the bandwidth pool, use the no form of this command.
gprs qos bandwidth-pool pool-name
no gprs qos bandwidth-pool pool-name
Syntax Description
Defaults
No bandwidth pools are configured.
Command Modes
Global configuration
Command History
Usage Guidelines
The CAC feature ensures that required network resources are available for real-time data traffic (such as voice, video, etc.). The CAC feature consists of two functions: maximum QoS authorization using CAC maximum QoS policies and bandwidth management.
The CAC bandwidth management function ensures that there is sufficient bandwidth for real-time PDP contexts during the PDP context activation and modification process.
The CAC feature uses user-defined bandwidth pools to negotiate and reserve bandwidth. In these pools, you define the total bandwidth allocated to that pool and then allocate a percentage of that bandwidth to each traffic class.
In the following example, bandwidth pool (pool A) has been created with 100000 kbps allocated to it. Additionally, a percentage of that 100000 kbps of bandwidth has been allocated to each traffic class, creating four "traffic class-based" bandwidth pools.
gprs bandwidth-pool Abandwidth 100000traffic-class conversational percent 40traffic-class streaming percent 30traffic-class interactive percent 20traffic-class background percent 10
Note
Once a bandwidth pool is allocated for a traffic class, it cannot be borrowed by the other sub pools allocated for the different traffic classes. The request is only admitted within the bandwidth pool to which the traffic class belongs.
Use the gprs qos bandwidth-pool command to create or modify a CAC bandwidth pool and apply the bandwidth pool to one or more APNs using the bandwidth-pool access point configuration command.
Examples
The following example creates a bandwidth pool named "pool a":
gprs qos bandwidth pool aRelated Commands
gprs qos cac-policy
To create or modify a Call Admission Control (CAC) maximum QoS policy that can be attached to one or more APNs, and enter CAC maximum QoS policy configuration mode, use the gprs qos cac-policy global configuration command. To return to the default value, use the no form of this command.
gprs qos cac-policy policy-name
no gprs qos cac-policy policy-name
Syntax Description
Defaults
No default behavior or values.
Command Modes
Global configuration
Command History
Usage Guidelines
The CAC feature on the GGSN ensures that required network resources are available for real-time data traffic such as voice and video. CAC is applied at the APN and consists of two functions: maximum QoS authorization and bandwidth management.
The CAC maximum QoS authorization function ensures that the QoS requested by a create PDP context does not exceed the maximum QoS configured within an APN. Using a CAC maximum QoS policy, you define certain QoS parameters within a policy and attach the policy to an APN. The CAC maximum QoS policy limits the QoS requested by the PDP during its creation and modification process.
Use the gprs qos cac-policy command to create or modify a CAC maximum QoS policy and apply the policy to an APN using the cac-policy access point configuration command.
Note
Once you have entered policy configuration mode using the gprs qos cac-policy command, you can configure the following QoS parameters in a policy and apply the policy to an APN:
•
•
•
•
•
•
•
Examples
The following example creates a CAC maximum QoS policy named "policy a":
gprs qos cac-policy aRelated Commands
gprs qos default-response requested
To specify that the GGSN sets its default QoS values in the response message exactly as requested in the create PDP context request message, use the gprs qos default-response requested global configuration command. To return to the default QoS, use the no form of this command.
gprs qos default-response requested
no gprs qos default-response requested
Syntax Description
This command has no arguments or keywords.
Defaults
Disabled. The GGSN sets its QoS default to the best-effort class.
Command Modes
Global configuration
Command History
Usage Guidelines
The gprs qos default-response requested command is only useful when canonical QoS is not configured on the GGSN. Canonical QoS is enabled using the gprs qos map canonical-qos command.
When canonical QoS is not enabled, and the gprs qos default-response requested command has not been configured on the GGSN, the GGSN always sets its QoS values to best-effort in the response message.
Examples
The following example enables the GGSN to set its QoS values in the response message according to the QoS values requested in the create PDP context request message:
gprs qos default-response requestedRelated Commands
gprs qos map canonical-qos
Enables mapping of GPRS QoS categories to a canonical QoS method that includes best-effort, normal, and premium QoS classes.
gprs qos map canonical-qos
To enable mapping of GPRS QoS categories to a canonical QoS method that includes best-effort, normal, and premium QoS classes, use the gprs qos map canonical-qos global configuration command. To disable canonical mapping, use the no form of this command.
gprs qos map canonical-qos
no gprs qos map canonical-qos
Syntax Description
This command has no arguments or keywords.
Defaults
Canonical QoS mapping is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the qprs qos map canonical-qos command to map GPRS QoS into the following canonical categories: best effort, normal, and premium.
Examples
The following example shows canonical QoS mapping enabled:
qos map canonical-qosRelated Commands
gprs qos map delay
To enable mapping of GPRS QoS categories to delay QoS classes, use the gprs qos map delay global configuration command. To disable delay mapping, use the no form of this command.
gprs qos map delay
no gprs qos map delay
Syntax Description
This command has no arguments or keywords.
Defaults
Disabled
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs qos map delay command to enable QoS delay mapping on the GGSN. To map the QoS delay classes (class 1, class 2, class 3, and best effort) to IP type of service (ToS) categories, use the gprs delay-qos map tos command.
Examples
The following example enables delay QoS mapping:
gprs qos map delayRelated Commands
gprs qos map umts
To enable UMTS QoS on the GGSN, use the gprs qos map umts global configuration command. To disable this mapping and return to the default QoS mapping, use the no form of this command.
gprs qos map umts
no gprs qos map umts
Syntax Description
This command has no arguments or keywords.
Defaults
UMTS QoS mapping is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs qos map umts command to enable UMTS QoS mapping.
Examples
The following example enables UMTS traffic QoS mapping:
gprs qos map umtsRelated Commands
gprs radius attribute chap-challenge
To specify that the CHAP challenge always be included in the Challenge Attribute field (and not in the Authenticator field) in an Access-Request to the Remote Access Dial-In User Service (RADIUS) server, use gprs radius attribute chap-challenge global configuration command. To disable, use the no form of this command.
gprs radius attribute chap-challenge
no gprs radius attribute chap-challenge
Syntax Description
This command has no arguments or keywords.
Defaults
If the CHAP challenge length is 16 bytes, it is sent in the Authenticator field of an Access-Request. If it is greater than 16 bytes, it is sent in the Challenge Attribute field.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs radius attribute chap-challenge command when configuring RADIUS security on the GGSN.
When the gprs radius attribute chap-challenge command is configured, the CHAP challenge is always sent in the Challenge Attribute field of an Access-Request to the RADIUS server and not in the Authenticator field. When the command is not configured, the CHAP challenge is sent in the Authenticator field unless the challenge exceeds 16 bytes, in which case, it is sent in the Challenge Attribute field of the Access-Request.
Examples
The following example configures the CHAP challenge to always be sent in an Access Request to the RADIUS server:
gprs radius msisdn first-bytegprs radius attribute session-timeout
To specify that the Session-Timeout (Attribute 27) field be included in a Remote Access Dial-In User Service (RADIUS) request, use the gprs radius attribute session-timeout command. To disable, use the no form of this command.
gprs radius attribute session-timeout
no gprs radius attribute session-timeout
Syntax Description
This command has no arguments or keywords.
Defaults
Attribute 27 is not included.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs radius attribute session-timeout command to configure the Session-Timeout (Attribute 27) field be included in a Remote Access Dial-In User Service (RADIUS) request.
The GGSN stores the attribute value received in Access-Accept packets sent by the AAA server and terminates the PDP context upon expiration of the time. You can configure the number of seconds the GGSN allows a session to be active before terminating the session at the global level (gprs gtp pdp-context timeout session command) and at the access-point level (gtp pdp-context timeout session command.
Examples
The following example configures Attribute 27 to always be sent in an Access Request to the RADIUS server:
gprs radius attribute session-timeoutRelated Commands
gprs radius msisdn first-byte
To specify that the first byte of the Mobile Stations International PSTN/ISDN (MSISDN) information element (IE) is included in a Remote Access Dial-In User Service (RADIUS) request, use the gprs radius msisdn first-byte global configuration command. To remove the first byte from the MSISDN IE in a RADIUS request, use the no form of this command.
gprs radius msisdn first-byte
no gprs radius msisdn first-byte
Syntax Description
This command has no arguments or keywords.
Defaults
The first byte is not included.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs radius msisdn first-byte command when configuring RADIUS security on the GGSN.
The first octet of an MSISDN IE using E.164 addressing is 91 in hexadecimal, that is 10010001. In this 91 code, the 1 is the extension bit, 001 is the international number, and 0001 indicates E.164 numbering.
Examples
The following example specifies that the first byte of the MSISDN IE is included in a RADIUS request:
gprs radius msisdn first-bytegprs redundancy
To enable GTP session redundancy (GTP-SR) on a GGSN, use the gprs redundancy global configuration command. To disable GTP-SR, use the no form of this command.
gprs redundancy
no gprs redundancy
Syntax Description
This command has no arguments or keywords.
Defaults
Disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs redundancy command to enable GTP-SR on a GGSN.
Cisco GGSN Release 5.1 and later supports Active/Standby, 1-to-1 inter-device GTP-SR. GTP-SR enables two GGSNs to appear as one network entity and ensures that continuous service is provided to mobile subscribers in the event one of the GGSNs fails.
In a GTP-SR implementation, the Active GGSN establishes and terminates PDP sessions and sends required stateful data to the Standby GGSN. To stay current on the states of active PDP sessions, the Standby GGSN receives the stateful data sent by the Active GGSN. As soon as the Standby GGSN detects that the Active GGSN has failed, it becomes active and assumes the responsibilities of the Active GGSN.
Before GTP-SR can be enabled on two redundant GGSNs, a GTP-SR inter-device infrastructure must be configured. For information on configuring a inter-device infrastructure, see the "Configuring GTP Session Redundancy" chapter of the Cisco GGSN Release 6.0 Configuration Guide.
Examples
The following example enables GTP-SR on a GGSN:
gprs redundancyRelated Commands
gprs redundancy charging sync-window cdr rec-seqnum
To configure the window size used to determine when the CDR record sequence number needs to be synchronized to the Standby GGSN, use the gprs redundancy charging sync-window cdr rec-seqnum global configuration command. To return to the default value, use the no form of this command.
gprs redundancy charging sync-window cdr rec-seqnum size
no gprs redundancy charging sync-window cdr rec-seqnum size
Syntax Description
size
Configures the window size used to determine when the CDR record sequence number needs to be synchronized.Valid range is 1 to 20.
Defaults
10
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs redundancy charging sync-window cdr rec-seqnum command to configure the window size used to determine when the record sequence number needs to be synchronized.
The record sequence number is used by the charging gateway to detect duplicate CDRs associated with a PDP context. To minimize the amount of data being synchronized to the Standby GGSN, the record sequence number is not synchronized each time a CDR is closed. Instead, a window threshold for the record sequence number is synchronized each time a CDR closes. The current value of the record sequence number and the record number last synchronized for a PDP context is checked, and if the difference is the value configured for the window size using the gprs redundancy charging sync-window cdr rec-seqnum global configuration command, the current record sequence number is synchronized to the Standby GGSN.
When a Standby GGSN becomes the Active GGSN, it starts from the last value synchronized, plus the window size.
Examples
The following example configures a window size of 15:
gprs redundancy charging sync-window cdr rec-seqnum 15Related Commands
gprs redundancy charging sync-window gtpp seqnum
To configure the window size used to determine when the GTP' sequence number needs to be synchronized to the Standby GGSN, use the gprs redundancy charging sync-window gtpp seqnum global configuration command. To return to the default value, use the no form of this command.
gprs redundancy charging sync-window gtpp seqnum size
no gprs redundancy charging sync-window gtpp seqnum size
Syntax Description
Defaults
10000
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs redundancy charging sync-window gtpp seqnum command to configure the window size used to determine when the GTP' sequence number needs to be synchronized.
The GTP' sequence number is used by the charging gateway to prevent the duplication of packets. The GGSN sends encoded CDRs associated with a PDP context in a GTP packet to the charging gateway. If the GTP packet is acknowledged by the charging gateway, it removes the packet from memory. If it is not acknowledged, it is retransmitted. The charging gateway cannot acknowledged GTP packets if the sequence number repeats.
To minimize the amount of data being synchronized to the Standby GGSN, the GTP' sequence number is not synchronized each time a CDR is closed. Instead, a window threshold for the GTP' sequence number is synchronized each time a CDR message is sent. The current value of the GTP' sequence number and the gtpp sequence number last synchronized for a PDP context is checked and if the difference is the value configured for the window size (using the gprs redundancy charging sync-window gtpp seqnum global configuration command), the current GTP prime sequence number is synchronized to the Standby GGSN.
When a Standby GGSN becomes the Active GGSN, it starts from the last value synchronized plus the window size.
Examples
The following example configures the window size for the GTP' sequence number synchronization to be 120:
gprs redundancy charging sync-window gtpp seqnum 120Related Commands
gprs service-aware
To enable service-aware billing on the GGSN, use the gprs service-aware global configuration command. To disable the support, use the no form of this command
gprs service-aware
no gprs service-aware
Syntax Description
This command has no arguments or keywords.
Defaults
Disabled.
Command Modes
Global configuration
Command History
12.3(14)YQ
This command was introduced.
12.3(14)YU
This command was integrated into Cisco IOS Release 12.3(14)YU.
Usage Guidelines
Use the gprs service-aware global configuration command to enable service-aware billing on the on the GGSN.
Note
Examples
The following configuration example enables service-aware billing on a GGSN:
gprs service-awareRelated Commands
gprs service-mode
To configure the global service-mode state of a GGSN, use the gprs service-mode global configuration command.
gprs service-mode {operational | maintenance}
Syntax Description
operational
Specifies that the service-mode state of the GGSN is operational.
maintenance
Specifies that the service-mode state of the GGSN is maintenance.
Defaults
Operational.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs service-mode command to place the global service-mode state of a GGSN in maintenance mode.
The GGSN service-mode function enables you to make configuration changes and test calls without affecting all active sessions on a GGSN. You can configure the service-mode state globally, for an access-point, and for the GGSN charging function. There are two service-mode states: operational and maintenance. The default is operational mode.
When a GGSN is placed in global maintenance mode, it rejects all new Create PDP Context requests. Therefore, no new PDP contexts are activated for an entire GGSN while it is in global maintenance mode.
Note
Examples
The following example places a GGSN in maintenance mode:
gprs service-mode maintenanceRelated Commands
gprs service-mode test imsi
To configure a test user for which you can create PDP contexts to test an APN configuration, use the gprs service-mode test imsi command. To remove the test user configuration, use the no form of this command.
gprs service-mode test imsi imsi-value
no gprs service-mode test imsi imsi-value
Syntax Description
imsi-value
International Mobile Subscriber Identity (IMSI) value for which PDP contexts are to be created.
Defaults
No test user is configured on the GGSN.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs service-mode test imsi command to configure a test user for which create PDP contexts will be created to test configurations.
Only one test user can be configured per GGSN.
Note
Examples
The following example creates a test user with the IMSI 211F111130000000:
gprs service-mode test imsi 211F111130000000Related Commands
gprs slb mode
To define the Cisco IOS SLB operation mode for GGSN-SLB messaging, use the gprs slb mode global configuration command.
gprs slb mode {dispatched | directed}
Syntax Description
dispatched
Specifies that the Cisco IOS SLB is operating in dispatched mode.
directed
Specifies that the Cisco IOS SLB is operating in directed server NAT mode.
Defaults
Dispatched
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs slb mode global configuration command to defined the Cisco IOS SLB operation mode. This command is used in conjunction with the gprs slb notify and the gprs slb vserver global configuration commands and is required for GGSN-SLB messaging when the Cisco IOS SLB is in directed server NAT mode.
If Cisco IOS SLB is in dispatched mode, the virtual server that forwarded the Create PDP Context request to the GGSN is known to the GGSN and the GGSn can send notification directly to that server. Therefore, only the gprs slb notify command is required to enable GGSN-SLB messaging on the GGSN.
If the Cisco IOS SLB is functioning in directed server NAT mode, the virtual server is not known to the GGSN.
Therefore, a list of virtual servers that the GGSN should notify must be defined on the GGSN using the gprs slb vserver global configuration command and the Cisco IOS SLB mode of operation must be defined using the gprs slb mode global configuration command. If the Cisco IOS SLB operation mode and virtual servers are not defined when the Cisco IOS SLB is functioning in directed server NAT mode, GGSN-SLB messaging is not enabled, even if the gprs slb notify command has been configured.
Note
Examples
The following example defines Cisco IOS SLB to be in directed server NAT mode:
gprs slb mode directedRelated Commands
gprs slb notify
To enable the GGSN to notify the Cisco IOS SLB when a specific condition occurs, use the gprs slb notify global configuration command. To disable GGSN-IOS SLB messaging, issue the no form of this command.
gprs slb notify {cac-failure}
no gprs slb notify {cac-failure}
Syntax Description
cac-failure
Specifies that the GGSN notify the Cisco IOS SLB when a UMTS QoS CAC or canonical QoS failure has caused a Create PDP Context request to be rejected.
Defaults
Disabled
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs slb notify command to enable GGSN-IOS SLB messaging.
The GGSN-IOS SLB messaging function enables you to configure a GGSN to notify the Cisco IOS SLB when a condition exists that affects a session forwarded by the Cisco IOS SLB, and to instruct the Cisco IOS SLB on how to react to the condition.
For example, if a Create PDP Context request is rejected by the GGSN because of a Call Admission Control (CAC) failure, the GGSN notifies the Cisco IOS SLB that a CAC-related failure has occurred and instructs the Cisco IOS SLB to reassign the session to another GGSN in the server farm.
Notification messages sent by the GGSN include the following information elements (IEs):
•
•
•
If the Cisco IOS SLB is functioning in dispatched mode, the virtual server that forwarded the create PDP context request to the GGSN is known to the GGSN and the GGSN can send notification directly to the server. Therefore, to enable GGSN-IOS SLB messaging when the Cisco IOS SLB is operating in dispatched mode only requires that the gprs slb notify command be configured.
If the Cisco IOS SLB is functioning in directed server NAT mode, the virtual server is not known to the GGSN. Therefore, a list of virtual servers that the GGSN should notify must be defined on the GGSN using the gprs slb vserver global configuration command and the Cisco IOS SLB mode of operation must be defined using the gprs slb mode global configuration command. If the Cisco IOS SLB operation mode and virtual servers are not defined when the Cisco IOS SLB is functioning in directed server NAT mode, GGSN-IOS SLB messaging is not enabled, even if the gprs slb notify command has been configured.
Examples
The following example configures the GGSN to notify the Cisco IOS SLB when a create PDP context request has been rejected because of a UMTS QoS CAC failure:
gprs slb notify cac-failureRelated Commands
gprs slb vserver
To configure the Cisco IOS SLB virtual servers to be notified by a GGSN when a condition configured using the gprs slb notify command exists and the Cisco IOS SLB is functioning in directed server NAT mode, use the gprs slb vserver global configuration command. To remove a virtual server from the list, use the no form of this command.
gprs slb vserver ip_address [next-hop ip ip-address [vrf name]]
no slb vserver ip_address [next-hop ip ip-address [vrf name]]
Syntax Description
Defaults
No virtual servers are defined.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs slb vserver global configuration command to defined a list of Cisco IOS SLB virtual servers to be notified by a GGSN when GGSN-SLB messaging is enabled and the Cisco IOS SLB is functioning in directed server NAT mode. This command is used in conjunction with the gprs slb notify and the gprs slb mode global configuration commands.
If the Cisco IOS SLB is functioning in directed server NAT mode, the virtual server is not known to the GGSN. Therefore, a list of virtual servers that the GGSN should notify must be defined on the GGSN using the gprs slb vserver global configuration command and the Cisco IOS SLB mode of operation must be defined using the gprs slb mode global configuration command. If the Cisco IOS SLB operation mode and virtual servers are not defined when the Cisco IOS SLB is functioning in directed server NAT mode, GGSN-SLB messaging is not enabled, even if the gprs slb notify command has been configured.
Note
Examples
Example 1
The following example adds a GTP server with the IP address 172.10.10.10 to the list of virtual servers to be notified by the GGSN:
gprs slb vserver 172.10.10.10Related Commands
gprs throughput interval
To configure the intervals at which the throughput data is collected for APNs, use the gprs throughput interval command. To return to the default value, use the no form of this command.
gprs throughput interval interval1 interval2
no gprs throughput interval interval1 interval2
Syntax Description
Defaults
No default behavior or values.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs throughput interval command to configure the intervals at which the GGSN will collect throughput data for APNs.
Examples
The following example configures the GGSN to collect throughput data every 5 minutes (300 seconds):
gprs throughput interval 300Related Commands
show gprs access-point throughput statistics
Displays throughput statistics for access points on a GGSN.
gprs umts-qos dscp unmodified
To specify that the subscriber datagram be forwarded through the GTP path without modifying its DSCP, use the gprs umts-qos dscp unmodified global configuration command. To remove this specification and enable the DSCP to be re-marked with the DSCP assigned to the traffic class during the PDP context creation, use the no form of this command.
gprs umts-qos dscp unmodified [up | down | all]
no gprs umts-qos dscp unmodified [up | down | all]
Syntax Description
Defaults
The DSCP in the subscriber datagram is re-marked with the DSCP assigned to the traffic class during the PDP context creation.
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs umts-qos dscp unmodified command to configure the GGSN to forward subscriber datagram DSCPs through the GTP path without modifying the DSCP.
Examples
The following example sets subscriber datagrams in the uplink GTP path to retain their DSCPs:
gprs umts-qos dscp unmodified upRelated Commands
gprs umts-qos map diffserv-phb
To assign a differentiated services code point (DSCP) to a DiffServ PHB group, use the gprs umts-qos map diffserv-phb global configuration command. To set the specified DSCP to the default DiffServ PHB group, use the no form of this command.
gprs umts-qos map diffserv-phb diffserv-phb-group [dscp1] [dscp2] [dscp3]
no gprs umts-qos map diffserv-phb
Syntax Description
Defaults
The default DSCP value associated with the PHB class is used.
Command Modes
Global configuration
Command History
Usage Guidelines
For the Assured Forwarding (AF) PHB group, you can specify up to three DSCP values for each drop precedence. The signalling, EF, and best-effort classes do not have drop precedence, so only the first DSCP value is used. If you enter a value for the dscp2 or dscp3 arguments for these classes, it is ignored.
Drop precedence indicates the order in which a packet will be dropped when there is congestion on the network.
Table 1 shows the default DSCP values for each PHB group.
Examples
The following example assigns a DSCP value of 31 to the EF class and three DSCP values to AF class2 of 51, 52, and 53:
gprs umts-qos map diffserv-phb ef-class 31gprs umts-qos map diffserv-phb af-class2 51 52 53Related Commands
gprs umts-qos map traffic-class
To specify a QoS mapping from the UMTS traffic classes to a differentiated services (DiffServ) per-hop behavior (PHB) group, use the gprs umts-qos map traffic-class global configuration command. To remove a QoS mapping and set the specified traffic class to the default mapping, use the no form of this command.
gprs umts-qos map traffic-class traffic-class diffserv-phb-group
no gprs umts-qos map traffic-class
Syntax Description
Defaults
You must enable UMTS QoS using the gprs qos map umts command before entering this command.
Note
The default mapping values for the UMTS traffic classes are as follows:
•
•
•
•
•
Command Modes
Global configuration
Command History
Usage Guidelines
Use the gprs umts-qos map traffic-class command to specify a mapping between various QoS UMTS traffic categories and the DiffServ PHB groups.
Examples
The following example specifies a QoS mapping from the UMTS traffic class conversational to the DiffServ PHB group af-class1:
gprs umts-qos map traffic-class conversational af1-classRelated Commands
gtp pdp-context single pdp-session
To configure the GGSN to delete the primary PDP context, and any associated secondary PDP contexts, of a hanging PDP session upon receiving a new create request from the same MS that shares the same IP address of the hanging PDP context, use the gtp pdp-context single pdp-session command. To return to the default value, use the no form of this command.
gtp pdp-context single pdp-session [mandatory]
[no] gtp pdp-context single pdp-session [mandatory]
Syntax Description
mandatory
Specifies that the primary PDP context and any associated secondary PDP contexts be deleted regardless of the RADIUS user profile configuration.
Defaults
Create PDP Context requests that share the IP address of an existing PDP context for the same MS are rejected.
Command Modes
Access-point configuration
Command History
Usage Guidelines
Use the gtp pdp-context single pdp-session command to configure the GGSN to delete the primary PDP context, and any associated secondary PDP contexts, of a hanging PDP session upon receiving a new create request from the same MS that shares the same IP address of the hanging PDP context.
A hanging PDP context is a PDP context on the GGSN whose corresponding PDP context on the SGSN has already been deleted for some reason.
When this condition occurs and the gtp pdp-context single pdp-session command is not configured, if on the same APN, the same MS sends a new Create PDP Context request that has a different NSAPI but has been assigned the same IP address used by the hanging PDP context, the GGSN rejects the new Create PDP Context request.
When the gtp pdp-context single pdp-session is configured on an APN, the single PDP session per MS feature is enabled and applies to all users for whom the "gtp-pdp-session=single-session" Cisco VSA has been defined in their RADIUS user profile. If the command is not configured, the feature is not enabled and does not apply to any user regardless of their RADIUS user profile configuration. If the command is configured with the mandatory keyword option specified, the feature is enabled and applies to all users on that APN regardless of their RADIUS user profile configuration.
Note
Examples
The following example configures the GGSN to delete the primary PDP context, and associated secondary PDP contexts, of a hanging PDP context when it receives a new create PDP context request that shares the same IP address:
gtp pdp-context single pdp-sessionRelated Commands
gtp pdp-context timeout idle
To specify the time, in seconds, that a GGSN allows a session to be idle at a particular access point before terminating the session, use the gtp pdp-context timeout idle access-point configuration command. To return to the default value, use the no form of this command.
gtp pdp-context timeout idle interval [uplink]
no gtp pdp-context timeout idle
Syntax Description
Defaults
259200 seconds (72 hours)
Command Modes
Access-point configuration
Command History
Usage Guidelines
The GGSN supports the RADIUS Idle-Timeout (Attribute 28) field. The GGSN stores the attribute 28 value if it is present in the access request packets sent by the AAA server. When a PDP context is idle for an amount of time that exceeds the session idle timeout duration, the GGSN terminates it.
The duration specified for the session idle timer applies to all PDP contexts of a session, however, a session idle timer is started for each PDP context. Therefore, the session idle timer is per-PDP, but the timer duration is per-session.
On the GGSN, the session idle timer can be configured globally and at the APN. The value configured at the APN level using the gtp pdp-context timeout idle access-point configuration command overrides the value configured globally using the gprs gtp pdp-context timeout idle global configuration command. The value configured in the user profile on the RADIUS server overrides the value configured at the APN.
Note
You can disable the session idle timer for a particular user by configuring 0 as the session idle time duration in the user profile on the RADIUS server. If a user is authenticated by RADIUS, the session idle time cannot be disabled.
Note
Note
Examples
The following example shows configuring the GGSN to wait 18000 seconds before ending an idle session:
gtp pdp-context timeout idle 18000Related Commands
gtp pdp-context timeout session
To specify the time, in seconds, that a GGSN allows a session to exist at a particular access point before terminating the session, use the gprs gtp pdp-context timeout session access-point configuration command. To return to the default value, use the no form of this command.
gtp pdp-context timeout session seconds
no gtp pdp-context timeout session seconds
Syntax Description
seconds
Time, in seconds, that the GGSN allows a session to exist at a particular access point. Specify a value between 30 and 4294967 seconds.
Defaults
Disabled
Command Modes
Access-point configuration
Command History
Usage Guidelines
When enabled using the gprs radius attribute session-timeout command, the GGSN supports the RADIUS Session-Timeout (Attribute 27). The GGSN stores the attribute timeout value received in access-accept packets sent by the AAA server and when the duration of a session exceeds the duration configured as absolute session timer, the GGSN terminates the session and all PDP contexts belonging to the session (those with the same IMSI or MS address).
Note
Note
On the GGSN, the absolute session timer can be configured globally and at the APN. The value configured at the APN level using the gtp pdp-context timeout session access-point configuration command overrides the value configured globally using the gprs gtp pdp-context timeout session global configuration command. The value configured in the user profile on the RADIUS server overrides the value configured at the APN.
Examples
The following example shows configuring the GGSN to wait 86400 seconds before ending a session:
gtp pdp-context timeout session 86400Related Commands
gtp response-message wait-accounting
To configure the GGSN to wait for a RADIUS accounting response before sending a create PDP context response to the SGSN, for create PDP context requests received at a particular APN, use the gtp response-message wait-accounting access-point configuration command. To configure the GGSN to send a create PDP context response to the SGSN after sending a RADIUS start accounting message to the RADIUS server (without waiting for a response from the RADIUS accounting server), use the no form of this command.
gtp response-message wait-accounting
no gtp response-message wait-accounting
Syntax Description
This command has no arguments or keywords.
Defaults
The GGSN sends a create PDP context response to the SGSN after sending a RADIUS start accounting message to the RADIUS accounting server. The GGSN does not wait for a RADIUS accounting response from the RADIUS accounting server.
Command Modes
Access-point configuration
Command History
Usage Guidelines
Use the gtp response-message wait-accounting command to configure the GGSN to wait for a RADIUS accounting response from the RADIUS accounting server, before sending a create PDP context response to the SGSN.
If the GGSN does not receive a response from the RADIUS accounting server when you have configured the gtp response-message wait-accounting command, then the GGSN rejects the PDP context request.
The GGSN supports configuration of RADIUS response message waiting at both the global and access-point configuration levels. You can minimize your configuration by specifying the configuration that you want to support across most APNs, at the global configuration level. Then, at the access-point configuration level, you can selectively modify the behavior that you want to support at a particular APN. Therefore, at the APN configuration level, you can override the global configuration of RADIUS response message waiting.
To configure the GGSN to wait for a RADIUS accounting response as the default behavior for all APNs, use the gprs gtp response-message wait-accounting global configuration command. To disable this behavior for a particular APN, use the no gtp response-message wait-accounting access-point configuration command.
To verify whether RADIUS response message waiting is enabled or disabled at an APN, you can use the show gprs access-point command and observe the value reported in the wait_accounting output field.
Examples
The following examples show only a partial configuration of the GGSN, to highlight those commands related to implementing RADIUS response message waiting. Additional configuration statements are required to complete a full configuration of the GGSN.
Example 1
The following example configures the GGSN to wait for an accounting response from the RADIUS server before sending a create PDP context response to the SGSN, for PDP context requests at access-point 1:
aaa new-model!aaa group server radius fooserver 10.2.3.4server 10.6.7.8!aaa authentication ppp foo group fooaaa authorization network default group radiusaaa accounting exec default start-stop group foo!gprs access-point-list gprsaccess-point 1access-mode non-transparentaccess-point-name www.pdn1.comaaa-group authentication foogtp response-message wait-accounting!radius-server host 10.2.3.4 auth-port 1645 acct-port 1646 non-standardradius-server host 10.6.7.8 auth-port 1645 acct-port 1646 non-standardradius-server key ggsntelExample 2
The following example globally configures the GGSN to wait for a RADIUS accounting response from the RADIUS server before sending a create PDP context response to the SGSN. The GGSN waits for a response for PDP context requests received across all access points, except access-point 1. RADIUS response message waiting has been overridden at access-point 1 using the no gtp response-message wait-accounting command:
aaa new-model!aaa group server radius fooserver 10.2.3.4server 10.6.7.8!aaa authentication ppp foo group fooaaa authorization network default group radiusaaa accounting exec default start-stop group foo!gprs access-point-list gprsaccess-point 1access-mode non-transparentaccess-point-name www.pdn1.comaaa-group authentication foono gtp response-message wait-accountingexitaccess-point 2access-mode non-transparentaccess-point-name www.pdn2.comaaa-group authentication foo!gprs gtp response-message wait-accounting!radius-server host 10.2.3.4 auth-port 1645 acct-port 1646 non-standardradius-server host 10.6.7.8 auth-port 1645 acct-port 1646 non-standardradius-server key ggsntelRelated Commands
interface
To specify the logical interface, by name, that the quota server will use to communicate with the Content Services Gateway (CSG), use the interface quota server configuration command. To remove the interface, use the no form of this command
interface interface-name
no interface interface-name
Syntax Description
Defaults
No default behavior or values.
Command Modes
Quota server configuration
Command History
12.3(14)YQ
This command was introduced.
12.3(14)YU
This command was integrated into Cisco IOS Release 12.3(14)YU.
Usage Guidelines
Use the interface quota server configuration mode command to specify the logical interface the quota server will use to communicate with the CSG.
We recommend that a loopback interface be used as the quota server interface.
If the path to the CSG is up, using the no form of this command will bring the path down. Therefore, ensure that you use the command carefully. It must be configured for proper quota server-to-CSG interworking.
Examples
The following configuration specifies the logical interface "loopback1" as the interface that the quota server will use to communicate with the CSG:
ggsn quota-server qs1interface loopback1Related Commands
.
ip probe path
To enable route probe support on an APN, use the ip probe path access-point configuration command. To return to the default, use the no form of this command.
ip probe path ip_address protocol udp [port port ttl ttl]
no ip probe path ip_address protocol udp [port port ttl ttl]
Syntax Description
Defaults
Disabled
Command Modes
Access-point configuration
Command History
Usage Guidelines
Use the ip probe path access-point configuration command to enable the GGSN to send a probe packet to a specific destination for each PDP context that is successfully established.
An example of how to use this feature is when a firewall load balancer (FWLB) is being used in the network. If the ip probe path command is configured, when a PDP context is established, the GGSN sends a probe packet the FWLB. This enables the FWLB to create an entry for the PDP context even if there is no upstream packet from the MS. Once an entry is created, the FWLB can forward any downstream packet from the network for the MS to the appropriate GGSN without depending on the MS to send the packet first.
Note
ip vrf forwarding
To associate a Virtual Private Network (VPN) routing/forwarding instance (VRF) with a Diameter peer, use the ip vrf forwarding Diameter peer configuration command. To remove the VRF configuration, use the no form of this command
ip vrf forwarding name
no ip vrf forwarding
Syntax Description
Defaults
The default is the global routing table.
Command Modes
Diameter peer configuration
Command History
12.3(14)YQ
This command was introduced.
12.3(14)YU
This command was integrated into Cisco IOS Release 12.3(14)YU.
Usage Guidelines
Use the ip vrf forwarding command to associate a VRF with a Diameter peer.
Examples
The following example shows how to link a VRF to Diameter peer "dcca1":
Router(config)# diameter peer dcca1Router(config-dia-peer)# ip vrf forwarding vpn1Related Commands
.
ip-access-group
To specify access permissions between an MS and a PDN through the GGSN at a particular access point, use the ip-access-group access-point configuration command. To disable the input access list, use the no form of this command.
ip-access-group access-list-number {in | out}
no ip-access-group access-list-number {in | out}
Syntax Description
Defaults
No access list is enforced.
Command Modes
Access-point configuration
Command History
Usage Guidelines
Use the ip-access-group command to specify an access list that indicates whether users are given or denied permission to access the mobile station from the PDN through the GGSN using a specified access point.
Examples
The following example grants access-list 101 inbound access to the mobile station from the PDN through the GGSN:
access-list 101 permit ip 10.0.0.2 0.255.255.255 anyinterface virtual-template 1ip unnumber loopback 1no ip directed-broadcastencapsulation gtpgprs access-point-list abc!gprs access-point-list abcaccess-point 1access-point-name gprs.somewhere.comdhcp-server 10.100.0.3ip-access-group 101 inexit!ip-address-pool
To specify a dynamic address allocation method using IP address pools for the current access point, use the ip-address-pool access-point configuration command. To return to the default value, use the no form of this command.
ip-address-pool {dhcp-proxy-client | radius-client | local pool-name | disable}
no ip-address-pool {dhcp-proxy-client | radius-client | local pool-name | disable}
Syntax Description
Defaults
The global setting specified with the gprs default ip-address-pool command is used. The default value for the global configuration command is that IP address pools are disabled.
Command Modes
Access-point configuration
Command History
Usage Guidelines
You can specify an IP allocation method for an access point in two ways:
•
•
If you specify dhcp-proxy-client as the method for allocating IP addresses, then you must configure a DHCP server for IP address allocation. You can do this at the global configuration level using the gprs default-dhcp server command, or at the access point level using the dhcp-server command.
If you specify radius-client as the method for allocating IP addresses, then you must configure a RADIUS server for IP address allocation, configure AAA on the GGSN, and configure AAA server groups globally on the GGSN or at the access point. For more information about configuring RADIUS on the GGSN, refer to the Usage Guidelines section for the aaa-group and gprs default aaa-group commands.
Note
Examples
The following example configures DHCP as the IP address pool allocation method for access-point 1 and specifies that the other access points use the global default, which is specified as RADIUS:
aaa new-model!aaa group server radius fooserver 10.2.3.4server 10.6.7.8aaa group server radius foo1server 10.10.0.1!aaa authentication ppp foo group fooaaa authentication ppp foo group foo1aaa authorization network default group radiusaaa accounting exec default start-stop group fooaaa accounting network foo1 start-stop group foo1!interface Loopback0ip address 10.88.0.1 255.255.255.255!interface virtual-template 1ip unnumber Loopback0no ip directed-broadcastencapsulation gtpgprs access-point-list abc!gprs access-point-list abcaccess-point 1access-point-name gprs.pdn1.comip address-pool dhcp-proxy-clientaggregate autodhcp-server 10.100.0.3dhcp-gateway-address 10.88.0.1exit!access-point 2access-point-name gprs.pdn2.comaccess-mode non-transparentaaa-group authentication fooexit!gprs default ip-address-pool radius-client!radius-server host 10.2.3.4 auth-port 1645 acct-port 1646 non-standardradius-server host 10.6.7.8 auth-port 1645 acct-port 1646 non-standardradius-server host 10.10.0.1 auth-port 1645 acct-port 1646 non-standardradius-server key ggsntelRelated Commands
limit duration
To specify as a trigger condition in a charging profile, the time duration limit that when exceeded causes the GGSN to collect upstream and downstream traffic byte counts and close and update the G-CDR for a particular PDP context, use the limit duration charging profile configuration command. To return to the default value, use the no form of this command.
limit duration number [reset]
no limit duration number [reset]
Syntax Description
Defaults
Disabled
Command Modes
Charging profile configuration
Command History
Usage Guidelines
Use the limit duration charging profile configuration command to specify the time limit, that when exceeded, causes the GGSN to collect upstream and downstream traffic byte counts and close and update the G-CDR for a PDP context.
For the box-level charging profile (profile 0 configured using the charging related global configuration commands), all triggers are reset by the expiration of another trigger. However, for charging profiles 1 through 15, the reset keyword option must be set for the limit duration and limit volume charging profile configuration commands for the expiration of any trigger to reset all other triggers.
If the reset keyword option is not specified when configuring the time trigger, the time trigger will not be reset when the volume trigger expires (limit volume command), but it will be reset when any other trigger expires.
Related Commands
.
limit sgsn-change
To specify as a trigger condition in a charging profile, the maximum number of SGSN changes that can occur before closing and updating the G-CDR for a particular PDP context, use the limit sgsn-change charging profile configuration command. To return to the default value, use the no form of this command.
limit sgsn-change number
no limit sgsn-change number
Syntax Description
Defaults
Disabled
Command Modes
Charging profile configuration
Command History
Usage Guidelines
A value of 0 means that a G-CDR is closed each time that a new SGSN begins handling the PDP context.
The command specifies the number of changes, not the number of SGSNs to be supported. The number of SGSNs supported is equal to 1 more than the change limit. For example, if the SGSN change limit is 2, the maximum number of SGSNs in the list before the GGSN closes the G-CDR is 3.
When you enable the gprs charging cdr-option no-partial-cdr-generation command, the GGSN creates any subsequent G-CDRs for the same PDP context request with the same fields in all G-CDRs and maintains sequence numbering.
If an SGSN change limit trigger is not configured when gprs charging cdr-option no-partial-cdr-generation command is configured, and a G-CDR is closed due to any other trigger (such as tariff times or QoS changes), the GGSN copies the last SGSN (the current SGSN) in the list in the new G-CDR. However, for charging releases prior to Release 4, by default, when the gprs charging cdr-option no-partial-cdr-generation command is configured and there is an SGSN change limit trigger configured either using the gprs charging container sgsn-change-limit global configuration or the limit sgsn-change charging profile configuration command, the CDR will not contain any SGSN address if it closed because of a non-SGSN-change trigger and there is no SGSN change. Therefore, to ensure that all CDR parameters are copied, including the SGSN list, specify the all keyword option when issuing the gprs charging cdr-option no-partial-cdr-generation.
Related Commands
.
limit volume
To specify as a trigger condition in a charging profile, the maximum number of bytes that the GGSN maintains across all containers for a particular PDP context before closing and updating the G-CDR, use the limit volume charging profile configuration command. To return to the default value, use the no form of this command.
limit volume threshold-value [reset]
no limit volume threshold-value [reset]
Syntax Description
Defaults
1,048,576 bytes (1 MB)
Command Modes
Charging profile configuration
Command History
Usage Guidelines
While a PDP context (mobile session) is active, charging events are generated based on various actions. One way that users can be charged is based on the amount of data transmitted between the PDN and the mobile station. Data volume is recorded in each of the containers of a G-CDR record. Service providers can use this recorded data volume to bill users by volume usage.
Use the limit volume charging profile configuration command to control the maximum amount of data volume that can be reported in each G-CDR from an active PDP context before the G-CDR is eligible for an update to the charging gateway for subsequent billing. The GGSN opens another partial G-CDR for that PDP context while it remains in session on the GGSN.
For example, consider that a volume threshold setting of 1 MB is configured on the GGSN. The GGSN opens a container in a G-CDR for a new PDP context. A trigger occurs for the PDP context, and at that time the GGSN has registered transmission of 500 KB of data for the PDP context. The trigger causes the GGSN to close the container for the PDP context, which has occurred before the volume limit is reached (500 KB of data transmitted, and 1 MB allowed).
As transmission for the PDP context continues, the GGSN opens a new container in the G-CDR. The GGSN now has up to 500 KB more data that can be processed for that PDP context before reaching the volume threshold limit for the G-CDR. When the volume threshold is reached across all containers for the PDP context (that is, the sum of all of the byte counts across all containers for the PDP context reaches 1 MB), the GGSN closes the G-CDR with a volume limit cause so that the G-CDR can be sent to the charging gateway. The GGSN opens another partial G-CDR for the PDP context while it remains in session.
For the box-level charging profile (profile 0 configured using the charging related global configuration commands), all triggers are reset by the expiration of another trigger. However, for charging profiles 1 through 15, the reset keyword option must be set for the limit duration and limit volume charging profile configuration commands for the expiration of any trigger to reset all other triggers. If the reset keyword is not specified when configuring the volume trigger, the volume trigger will not be reset when the time trigger expires (limit duration command), but it will be reset when any other trigger expires.
Related Commands
.
match flow pdp
To specify PDP flows as the match criterion in a class map, use the match flow pdp class map configuration command. To remove PDP flow as a match criterion, use the no form of this command.
match flow pdp
no match flow pdp
Syntax Description
This command has no arguments or keywords.
Defaults
No default behavior or values.
Command Modes
Class map configuration
Command History
Usage Guidelines
The match flow pdp class map configuration command enables the ability to configure session-based policing (per-PDP policing) for downlink traffic on a GGSN.
Note
Note
Note
To configure the Per-PDP policing feature on a GGSN, you must complete the following tasks:
1.
GGSN(config)# class-map class-pdpGGSN(config-cmap)# Match flow pdpGGSN(config-cmap)# exit2.
GGSN(config)# policy-map policy-gprsGGSN(config-pmap)# class class-pdp3.
GGSN(config-pmap-c)# police rate pdp [burst bytes] [peak-rate pdp [peak-burst bytes]] conform-action action exceed-action action [violate-action action]GGSN(config-pmap-c)# exitGGSN(config-pmap)# exit4.
GGSN(config)# access-point 1GGSN(access-point-config) service-policy in policy-gprsExamples
The following example specifies PDP flows as the match criterion in a class map named "class-pdp":
class-map class-pdpmatch flow pdpRelated Commands
police rate
Configures traffic policing using the police rate.
service-policy
Attaches a service policy to an APN, to be used as the service policy for PDP flows of that APN.
maximum delay-class
To define in a Call Admission Control (CAC) maximum QoS policy, the maximum delay class for R97/R98 QoS that can be accepted at an APN, use the maximum delay-class CAC maximum QoS policy configuration command. To return to the default value, use the no form of this command.
maximum delay-class value [reject]
no maximum delay-class value [reject]
Syntax Description
Defaults
PDP contexts for which the maximum delay-class is higher than the configured value are downgraded to the configured value.
Command Modes
CAC maximum QoS policy configuration
Command History
Usage Guidelines
Use the maximum delay-class CAC maximum QoS policy configuration command to specify the maximum delay class that can be accepted at an APN.
By default, PDP contexts for which the maximum delay-class is higher than the configured value are downgraded to the configured value.
If the reject keyword has been specified, if the maximum delay class requested is higher than the configured delay class, the create PDP context is rejected.
If the reject keyword is not specified and the delay class in a create or update PDP context request is greater than the configured value, the requested delay class is downgraded to the configured value.
Examples
The following example defines 3 as the maximum delay class for GPRS QoS that can be accepted at an APN:
maximum delay-class 3Related Commands
maximum pdp-context
To specify in a Call Admission Control maximum QoS policy, the maximum number of PDP contexts that can be created for a particular APN, use the maximum pdp-context CAC maximum QoS policy configuration command. To return to the default value, use the no form of this command.
maximum pdp-context number1 [threshold number2]
no maximum pdp-context number1 [threshold number2]
Syntax Description
Defaults
No default behavior or values.
Command Modes
CAC maximum QoS policy configuration
Command History
Usage Guidelines
Use the maximum pdp-context CAC maximum QoS policy configuration command to configure the maximum number of PDP contexts that can be created for a particular APN.
The maximum number of PDP contexts defined for an APN using the maximum pdp-context command cannot exceed the maximum number of PDP contexts defined by the gprs maximum-pdp-context-allowed global configuration command.
When the optional threshold keyword is specified, when the total number of PDP contexts exceeds the configured number, only PDP contexts with Allocation/Retention Priority 1 are accepted. Create PDP contexts with other priorities (2/3) are rejected. If the optional threshold keyword is not specified, when the total number of PDP contexts reaches the configured maximum number, all subsequent create PDP contexts are rejected.
The maximum pdp-context command configuration is checked before all other QoS parameters defined in a policy: maximum bit rate, guaranteed bit rate, highest traffic class, highest traffic handling priority, highest delay class, and highest peak throughput class.
Examples
In the following example, 15000 is specified as the maximum number of PDP contexts that can be created for a particular APN:
maximum pdp-context 15000Related Commands
maximum peak-throughput
To define in a Call Admission Control (CAC) maximum QoS policy, the maximum peak throughput for R97/R98 QoS that can be accepted at an APN, use the maximum peak-throughput CAC maximum QoS policy configuration command. To return to the default value, use the no form of this command.
maximum peak-throughput value [reject]
no maximum peak-throughput value [reject]
Syntax Description
Defaults
PDP contexts for which the peak throughput is higher than the configured value are downgraded to the configured value.
Command Modes
CAC maximum QoS policy configuration
Command History
Usage Guidelines
Use the maximum peak-throughput CAC maximum QoS policy configuration command to specify the maximum peak throughput that can be accepted at an APN.
By default, PDP contexts for which the peak throughput is higher than the configured value are downgraded to the configured value.
If the reject keyword has been specified, if the maximum peak throughput requested is higher than the configured peak throughput, the create PDP context is rejected.
If the reject keyword is not specified and the peak throughput in a create or update PDP context request is greater than the configured value, the requested peak throughput is downgraded to the configured value.
Examples
The following example defines 7 as the maximum peak-throughput GPRS QoS that can be accepted at an APN:
maximum peak-throughput 7Related Commands
maximum traffic-class
To define in a Call Admission Control (CAC) maximum QoS policy, the highest traffic class that can be accepted at an APN, use the maximum traffic-class CAC maximum QoS policy configuration command. To return to the default value, use the no form of this command.
maximum traffic-class traffic-class-name [priority value]
no maximum traffic-class traffic-class-name [priority value]
Syntax Description
Defaults
All traffic classes are accepted.
Command Modes
CAC maximum QoS policy configuration
Command History
Usage Guidelines
Use the maximum traffic-class CAC maximum QoS policy configuration command to define the highest traffic class that can be accepted at an APN. If the traffic class requested in a create PDP context request is higher than the configured class, the request is rejected.
The GGSN does not downgrade the traffic class of a PDP context unless the highest traffic class configured is changed after a PDP context is created and the GGSN receives an update PDP context request with a traffic class higher than the newly configured value. If this condition occurs, the GGSN downgrades the traffic class to the value of the newly configured maximum traffic class.
By default, all traffic classes are accepted.
Use the optional priority keyword to define the highest traffic handling priority for the interactive traffic class. If the requested traffic handling priority exceeds the highest one, it will be downgraded to the configured one. If the interactive traffic class is configured without the priority keyword option, then PDPs with any traffic handling priority are allowed. If the traffic class is not interactive, the priority keyword is ignored.
Examples
The following example configures streaming as the highest traffic class accepted at an APN:
maximum traffic-class streamingThe following example configures interactive as the highest traffic class accepted at an APN:
maximum traffic-class interactiveThe following example configures interactive as the highest traffic class with traffic handling priority 2 accepted at an APN:
maximum traffic-class interactive priority 2Related Commands
mbr traffic-class
To define in a Call Admission Control (CAC) maximum QoS policy, the maximum bit rate (MBR) that can be allowed for each traffic class, use the mbr traffic-class CAC maximum QoS policy configuration command. To return to the default value, use the no form of this command.
mbr traffic-class traffic-class-name bitrate {uplink | downlink} [reject]
no mbr traffic-class traffic-class-name bitrate {uplink | downlink} [reject]
Syntax Description
Defaults
Any MBR is accepted.
Command Modes
CAC maximum QoS policy configuration
Command History
Usage Guidelines
Use the mbr traffic-class CAC maximum QoS policy configuration command to define the highest MBR that can be accepted for real-time traffic on an APN.
When the reject optional keyword is specified, if the requested MBR exceeds the configured value, create PDP contexts are rejected. If the reject keyword is not specified, the MBR is downgraded to the configured value.
If the reject keyword is not specified and the MBR in a create or update PDP context request is greater than the configured value, the requested MBR is downgraded to the configured value.
Examples
The following example defines 1000 kbps as the uplink MBR supported and 2000 kbps as the maximum downlink MBR:
mbr traffic-class interactive 1000 uplinkmbr traffic-class interactive 1000 downlinkRelated Commands
