FastPCRF settings [Документация VAS Experts]

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision
Previous revision
en:dpi:bras_bng:radius_integration:radius_accounting:setup:start [2023/10/13 13:15] – ↷ Page moved from en:dpi:bras_bng:opt_bras_l3:radius_accounting:setup:start to en:dpi:bras_bng:radius_integration:radius_accounting:setup:start elena.krasnobryzhen:dpi:bras_bng:radius_integration:radius_accounting:setup:start [Unknown date] (current) – removed - external edit (Unknown date) 127.0.0.1
Line 1: Line 1:
-==== FastPCRF settings ==== 
-{{indexmenu_n>3}} 
- 
-Radius Accounting packages are sent by default to 1813 Radius server port. 
-You can specify a different port globally for all the radius servers or specifically for a given radius server: 
-  * ''radius_acct_port'' option defined in the fastpcrf.conf configuration file specifies the default accounting port 
-  * ''acct_port'' option is applied for a given radius server: 
-''radius_server=secret@192.168.200.1%eth1:1812;acct_port=34567'' 
- 
-You can also specify a secret for Radius accounting for each server separately; this can be done using the 
-''acct_secret'' option of ''radius_server'' parameter: 
- 
-''radius_server=qwerty@192.168.200.1%eth1:1812;acct_port=34567;acct_secret=123456'' 
- 
-The ''qwerty'' secret will be used for authorization requests in the example above, and ''123456'' will be used for the accounting ones. 
- 
-The following additional parameters related to the Radius Accounting can be specified in the fastpcrf.conf: 
- 
-♦ ''radius_max_acct_connect_count'' – the maximum number of connections used to transfer the accounting data, 
-the default value is 2. Each connection can be used to handle no more than 256 requests simultaneously. 
-This parameter limits the maximum Radius server load. 
- 
-♦ ''radius_acct_interim_interval'' – the time interval, in seconds, for sending the interim accounting. 
-The default value is 1800 seconds. The minimum allowed value is 60 seconds.  
-It should be noted that the time interval for sending interim statistics can be specified for each user 
-separately using the Acct-Interim-Interval attribute containing within the Access-Accept/Access-Reject response to the Access-Request authorization request. Value 0 means that interim statistics will not be sent. 
- 
-{{anchor:radius_acct_idle_timeout}} 
-♦ ''radius_acct_idle_timeout'' - [is added in the VAS Experts DPI 7.3] the time interval, in seconds, during which the data is unchanged. If within this time interval, there was no change in accounting data being sent by the fastdpi,then the accounting-session is considered to be closed since the idle time-out is exceeded. The default value is 1800 seconds. It should be noted that it is meaningless to set the parameter value less than ''netflow_timeout'' one. 
- 
-{{anchor:acct_check_idle_mode}} 
-♦ ''acct_check_idle_mode'' - [is added in the VAS Experts DPI 7.4] The stop event control mode, is used to control the event when the data will be changed. 
-Valid values are the following: 
-    * 0 - within the ''radius_acct_idle_timeout'' there is no changes in the incoming and outgoing traffic (this is the default value) 
-    * 1 - within the ''radius_acct_idle_timeout'' there is no changes only in the outgoing traffic (there is no subscriber's activity). 
-The ''acct_check_idle_mode=1'' mode allows to detect when the accounting session terminates even if the subscriber IP address is flooded with superfluous requests (for example, DDoS-attack). 
- 
-♦ ''acct_start_point'' - [is added in the VAS Experts DPI 7.4] The starting point of the accounting statistics: 
-    * 0 - since the last Stop (default value). This is the most accurate statistics; 
-    * 1 - since the Start event. 
-When the ''acct_start_point=0'' is used within the ''acct_check_idle_mode=1'' mode all the "parasite" traffic being accumulated since the last Stop event will fall into new accounting session.\\ 
-When the ''acct_start_point=1'' is used instead all the "parasite" traffic being accumulated since the last Stop event will be ignored. 
-Also, due to the time delays between accounting "activating" and the actual netflow statistics gathering, the outgoing traffic statistics may be slightly distorted (downwards).