Dynamic Routing is a module for selecting (based on multiple criteria) the the best gateway/destination to be used for delivering a certain call. Least Cost Routing (LCR) is a special case of dynamic routing - when the rules are ordered based on costs. Dynamic Routing comes with many features regarding routing rule selection:

prefix based
caller/group based
time based
priority based

, processing :

stripping and prefixing
default rules
inbound and outbound processing
script route triggering

and failure handling:

serial forking
weight based GW selection
random GW selection

1.2. Features

The dynamic routing implementation for Kamailio is designed with the following properties:

routing info (destinations, rules, groups) are stored in a database and loaded into memory at start up time; reload at runtime via MI command
load balancing or random selection of the destinations (from a given set)
able to handle large volume of routing info (300K of rules) with minimal speed/time and memory consumption penalties
script integration - Pseudo variables support in functions; scripting route triggering when rules are matched
bidirectional behavior - inbound and outbound processing (strip and prefixing when sending and receiving from a destination/GW)

1.3. Performance

There were several tests performed regarding the performance of the module when dealing with a large number of routing rules.

The tests were performed with a set of 383000 rules and to values were measured:

time to load from DB
used shared memory

The time to load was varying between 4 seconds and 8 seconds, depending of the caching of the DB client - the first load was the slowest (as the DB query hits the disk drive); the following are faster as data is already cached in the DB client. So technically speaking, the time to load (without the time to query which is DB type dependent) is ~4 seconds

After loading the data into shared memory ~ 96M of memory were used exclusively for the DR data.

1.4. Routing Rule Definition

Dynamic routing rules are stored in a database, in four tables:

one for storing the gateway definitions
one for storing the routing rule definitions
one for storing the users mappings over groups
one for storing a list of gateways, so you don't have to enter all the elements every time you need it

1.4.1. Gateway Addresses

Default name for the table storing gateway addresses is “dr_gateways”. Gateway addresses are stored in a separate table because of need to access them independent of Dynamic Routing processing (e.g., adding/ removing gateway PRI prefix before/after performing other operation -- receiving/relaying to gateway).

Table 1.1. Definition of table dr_gateways

Column name	Type	Default value	Description
gwid	integer	auto increment	unique identifier for GW
type	unsigned int	0	type/class of GW
address	varchar(128)		address of the gateway
strip	unsigned int	0	no of digits to strip
pri_prefix	varchar(255)		PRI prefix of the gateway
description	varchar(128)		description of the gateway

Once a rule is matched, the STRIP number of digits are removed from the username part of the RURI and then the PRI prefix has to be added to the request URI before forwarding the call to the gateway.

Table 1.2. Sample dr_gateways records

gwid	type	address	strip	pri_prefix	description
1	10	10.10.10.10:5080	0	2222	Gateway 1
2	10	10.10.10.10	2	3333	Gateway 2
3	20	10.10.10.11	0		Gateway 3

1.4.2. Destination/GW lists

For each rule, you can set a list of destinations to be used. The list is comma or pipe separated enumeration of the destinations. The module will use (one by one) each destination from the list (in the given order).

Also the module allows the usage of groups in the destination lists. A group of destinations is delimited by semi-colon char. inside the whole destination list ( like: 2,4;5,78,23;4;7;2 ). The destinations from within a group may be act differently (like load-balancing, random selection, etc), depending of the “sort_order” module parameter - more about this is available under the module paramters section.

1.4.3. Routing Rules

Default name for the table storing rule definitions is “dr_rules”.

Table 1.3. Definition of dr_rules table

Column name	Type	Default	Description
ruleid	integer	auto	UID of the rule
groupid	varchar(255)		list of routing group IDs
prefix	varchar(64)		destination prefix for this route
timerec	varchar(255)		time recurrence for this rule
priority	integer	0	priority of the rule
routeid	integer	0	route block to be execute
gwlist	varchar(255)		the list with GWs to be used
description	varchar(128)		description of this rule

groupid column

Each user must be member of only one routing group. It must be specified in user's profile.
prefix column

Destination URI must start with prefix value to match the rule.

time rec column

A date-time expression that defines the time recurrence to match for current rule. Time recurrences are based closely on the specification of recurring intervals of time in the Internet Calendaring and Scheduling Core Object Specification (calendar COS), RFC 2445. The set of attributes used in routing rule specification is subset of time recurrence attributes used by Call Processing Language (CPL). These attributes are (extracted from CPL draft 09):

Table 1.4. Time recurrence attributes

Attribute	Description
dastard	Start of interval (RFC 2445 DATE-TIME)
duration	Length of interval (RFC 2445 DURATION)
freq	Frequency of recurrence (secondly,minutely,hourly, daily,weekly, monthly, or yearly).
until	bound of recurrence (RFC 2445 DATE-TIME)
interval	How often the recurrence repeats
byday	List of days of the week
bymonthday	List of days of the month
byyearday	List of days of the year
byweekno	List of weeks of the year
bymonth	List of months of the year

When an attribute is not specified, the corresponding place must be left empty, whenever another attribute that follows in the list has to be specified.

Detailed description of time recurrence attributes:

dtstart - specifies the beginning of the first period.
duration - specifies the duration of the period. For a recurring interval, the “duration” parameter MUST be small enough such that subsequent intervals do not overlap. For non-recurring intervals, durations of any positive length are permitted, zero-length duration means “forever”. Negative-length durations are not allowed.
freq - takes one of the following values: “daily”, to specify repeating periods based on an interval of a day or more; “weekly”, to specify repeating periods based on an interval of a week or more; “monthly”, to specify repeating periods based on an interval of a month or more; and “yearly”, to specify repeating periods based on an interval of a year or more. These values are not case-sensitive.
until - defines an iCalendar COS DATE or DATE-TIME value which bounds the recurrence rule in an inclusive manner. If the value specified by “until” is synchronized with the specified recurrence, this date or date-time becomes the last instance of the recurrence. If not present, the recurrence is considered to repeat forever.
interval - contains a positive integer representing how often the recurrence rule repeats. The default value is “1”, meaning every day for a “daily” rule, every week for a “weekly” rule, every month for a “monthly” rule and every year for a “yearly” rule.
interval - contains a positive integer representing how often the recurrence rule repeats. The default value is “1”, meaning every day for a “daily” rule, every week for a “weekly” rule, every month for a “monthly” rule and every year for a “yearly” rule.
byday - specifies a comma-separated list of days of the week. “MO” indicates Monday; “TU” indicates Tuesday; “WE” indicates Wednesday; “TH” indicates Thursday; “FR” indicates Friday; “SA” indicates Saturday; “SU” indicates Sunday. These values are not case-sensitive.

Each “byday” value can also be preceded by a positive (+n) or negative (-n) integer. If present, this indicates the nth occurrence of the specific day within the “monthly” or “yearly” recurrence. For example, within a “monthly” rule, +1MO (or simply 1MO) represents the first Monday within the month, whereas -1MO represents the last Monday of the month. If an integer modifier is not present, it means all days of this type within the specified frequency. For example, within a “monthly” rule, MO represents all Mondays within the month.
bymonthday - parameter specifies a comma-separated list of days of the month. Valid values are 1 to 31 or -31 to -1. For example, -10 represents the tenth to the last day of the month.
byyearday - specifies a comma-separated list of days of the year. Valid values are 1 to 366 or -366 to -1. For example, -1 represents the last day of the year (December 31st) and -306 represents the 306th to the last day of the year (March 1st).
byweekno - specifies a comma-separated list of ordinals specifying weeks of the year. Valid values are 1 to 53 or -53 to -1.
bymonth - parameter specifies a comma-separated list of months of the year. Valid values are 1 to 12.

A recurrence is specified by including the “freq” parameter, which indicates the type of recurrence rule. Parameters other than “dtstart” and “duration” SHOULD NOT be specified unless “freq” is present.

If byxxx parameter values are found which are beyond the available scope (ie, bymonthday=“30” in February), they are simply ignored.

Byxxx parameters modify the recurrence in some manner. Byxxx rule parts for a period of time which is the same or greater than the frequency generally reduce or limit the number of occurrences of the recurrence generated. For example, freq=“daily” bymonth=“1” reduces the number of recurrence instances from all days (if the “bymonth” parameter is not present) to all days in January. Byxxx parameters for a period of time less than the frequency generally increase or expand the number of occurrences of the recurrence. For example, freq=“yearly” bymonth=“1,2” increases the number of days within the yearly recurrence set from 1 (if “bymonth” parameter is not present) to 2.

If multiple Byxxx parameters are specified, then after evaluating the specified “freq” and “interval” parameters, the Byxxx parameters are applied to the current set of evaluated occurrences in the following order: “bymonth”, “byweekno”, “byyearday”, “bymonthday”, “byday”; then “until” is evaluated.

Here is an example of evaluating multiple Byxxx parameters.

dtstart=“19970105T083000” duration=“10M” freq=“yearly” interval=“2” bymonth=“1” byday=“SU”

First, the interval=“2” would be applied to freq=“yearly” to arrive at “every other year” . Then, bymonth=“1” would be applied to arrive at “every January, every other year”. Then, byday=“SU” would be applied to arrive at “every Sunday in January, every other year, from 8:30 to 8:40 ”. The appropriate minutes and hours have been retrieved from the “dtstart” and “duration” parameters.

priority column

If many rules are eligible, choose the one with highest priority.
routeid column

If different than 0, then execute the route with the specified ID. That is, a route which can be used to perform custom operations on message. At this route, no modification is performed at signaling level.
gwlist column

A comma separated list of gateway identifiers corresponding to a row in table “dr_gateways”. The first gateway is tried first and if failure the second one, and so one. If no gateway is left a negative response is sent back to the user.

Routing Rules Examples

Table 1.5. Sample dr_rules records

ruleid	group	prefix	timerec	priority	routeid	gwlist	description
1	6	0049	20040101T083000\|10H\|weekly\|\|\|MO,TU,WE,TH,FR	5	23	1,2	Rule 1
2	8	0049	20040101T083000	0	0	1,2	Rule 2
3	7,8,9	0049	20040101T083000	0	0	3	Rule 3

(The time recurrence for first rule is: “20040101T083000|10H|weekly|||MO,TU,WE,TH,FR”)

1.5. Routing Rule Processing

The module can be used to find out which is the best gateway to use for new calls terminated to PSTN. The algorithm to select the rule is as follows:

the module discovers the routing group of the originating user. This step is skipped if a routing group is passed from the script as parameter.
once the group is known, in the subset of the rules for this group the module looks for the one that matches the destination based on "prefix" column. The set of rules with the longest prefix is chosen. If no digit from the prefix matches, the default rules are used (rules with no prefix)
within the set of rules is applied the time criteria, and the rule which has the highest priority and matches the time criteria is selected to drive the routing.
Once found the rule, it may contain a route ID to execute. If a certain flag is set, then the processing is stopped after executing the route block.
The rule must contain a gateway chain. The module will execute serial forking for each address in chain. The next address in chain is used only if the previously has failed.
With the right gateway address found, the prefix (PRI) of the gateway is added to the request URI and then the request is forwarded.

If no rule is found to match the selection criteria an default action must be taken (e.g., error response sent back). If the gateway in the chain has no prefix the request is forwarded without adding any prefix to the request URI.

2. Dependencies

2.1. Kamailio Modules

The following modules must be loaded before this module:

a database module.

2.2. External Libraries or Applications

none.

3. Exported Parameters

3.1. `db_url`(str)

The database url.

Default value is “NULL”.

Example 1.1. Set db_url parameter

...
modparam("drouting", "db_url", 
	"mysql://openser:openserrw@localhost/openser")
...

3.2. `drd_table`(str)

The name of the db table storing gateway addresses.

Default value is “dr_gateways”.

Example 1.2. Set drd_table parameter

...
modparam("drouting", "drd_table", "dr_gateways")
...

3.3. `drr_table`(str)

The name of the db table storing routing rules.

Default value is “dr_rules”.

Example 1.3. Set drr_table parameter

...
modparam("drouting", "drr_table", "rules")
...

3.4. `drg_table`(str)

The name of the db table storing groups.

Default value is “dr_groups”.

Example 1.4. Set drg_table parameter

...
modparam("drouting", "drg_table", "groups")
...

3.5. `drl_table`(str)

The name of the db table storing definitions of destination lists (to be used directly by the routing rules). You will have a identifier to a group of gateways instead of having all the members of the group as a individual elements. Very useful to reuse a list of gateways in different places.

Default value is “dr_gw_lists”.

Example 1.5. Set drl_table parameter

...
modparam("drouting", "drl_table", "my_gw_lists")
...

3.6. `sort_order` (int)

Defines how the destination list should be processed (ordering of the elements). Possible modes are

0 - destination groups are ignored and all the destinations are tried in the given order; Ex: list 1,2;3,4,5;6 will lead to usage as 1,2,3,4,5,6
1 - the destinations from each group are randomly arranged (only the two elements are randomly selected); groups do maintain their order (as given); the resulting list is used (with all the defined destinations). Ex: 1,2;3,4,5;6 -> randomizer -> (A) 2,1;4,3,5;6 -> usage 2,1,4,3,5,6 (B) 1,2;3,5,4;6 -> usage 1,2,3,5,4,6
2 - from each destination group, only a single destination is randomly selected; groups do maintain their order (as given); Ex: 1,2;3,4,5;6 -> randomizer -> (A) 2;4;6 -> usage 2,4,6 (B) 1;5;6 -> usage 1,5,6

Default value is “0”.

Example 1.6. Set sort_order parameter

...
modparam("drouting", "sort_order", 2)
...

3.7. `ruri_avp` (str)

The name of the avp for storing Request URIs to be later used (alternative destiantions for the current one).

Default value is “NULL”.

Example 1.7. Set ruri_avp parameter

...
modparam("drouting", "ruri_avp", '$avp(dr_ruri)')
modparam("drouting", "ruri_avp", '$avp(i:33)')
...

3.8. `attrs_avp` (str)

The name of the avp for storing the attribute of the current selected destination - once a new destination is selected (via the use_next_gw() function), the AVP will be updated with the attrs of the new used destination.

Default value is “NULL”.

Example 1.8. Set attrs_avp parameter

...
modparam("drouting", "attrs_avp", '$avp(dr_attrs)')
modparam("drouting", "atrrs_avp", '$avp(i:67)')
...

3.9. `use_domain` (int)

Flag to configure whether to use domain match when querying database for user's routing group.

Default value is “1”.

Example 1.9. Set use_domain parameter

...
modparam("drouting", "use_domain", 0)
...

3.10. `drg_user_col` (str)

The name of the column in group db table where the username is stored.

Default value is “username”.

Example 1.10. Set drg_user_col parameter

...
modparam("drouting", "drg_user_col", "user")
...

3.11. `drg_domain_col` (str)

The name of the column in group db table where the domain is stored.

Default value is “domain”.

Example 1.11. Set drg_domain_col parameter

...
modparam("drouting", "drg_domain_col", "host")
...

3.12. `drg_grpid_col` (str)

The name of the column in group db table where the group id is stored.

Default value is “groupid”.

Example 1.12. Set drg_grpid_col parameter

...
modparam("drouting", "drg_grpid_col", "grpid")
...

3.13. `fetch_rows` (int)

The number of rows that should be fetched from the result of a query in rules db table.

Default value is “2000”.

Example 1.13. Set fetch_rows parameter

...
modparam("drouting", "fetch_rows", 1500)
...

3.14. `force_dns` (int)

Force DNS resolving of GW/destination names (if not IPs) during startup. If not enabled, the GW name will be blindly used during routing.

Default value is “1 (enabled)”.

Example 1.14. Set force_dns parameter

...
modparam("drouting", "force_dns", 0)
...

4. Exported Functions

4.1. `do_routing("[groupID]")`

Function to trigger routing of the message according to the rules in the database table and the configured parameters.

This function can be used from REQUEST_ROUTE and FAILURE_ROUTE.

The module can take one optional parameter: the routing group the caller belongs to - this may be a static numerical value or an AVP specification. If none specified, the function will automatically try to query the dr_group table to get this information.

Example 1.15. do_routing usage

...
do_routing();
...
do_routing("0");
...
do_routing("$avp(i:10)");

4.2. `use_next_gw()/next_routing()`

The function takes the next available destination (set by do_routing, as alternative destinations) and push it into RURI. Note that the function just sets the RURI (nothing more).

If a new RURI is set, the used destination is removed from the pending set of alternative destinations.

This function can be used from REQUEST_ROUTE and FAILURE_ROUTE.

The function returns true only if a new RURI was set. False is returned is no other alternative destinations are found or in case of internal processing error.

Example 1.16. use_next_gw usage

...
if (use_next_gw()) {
	t_relay();
	exit;
}
...

4.3. `goes_to_gw([type])`

Function returns true if the destination of the current request (destination URI or Request URI) points (as IP) to one of the gateways. There no DNS lookups done if the domain part of the URI is not an IP.

This function does not change anything in the message.

This function can be used from REQUEST_ROUTE and FAILURE_ROUTE.

The function can take one optional parameter:

type (optional) - GW/destination type to be checked

Example 1.17. goes_to_gw usage

...
if (goes_to_gw("1")) {
	sl_send_reply("403","Forbidden");
	exit;
}
...

4.4. `is_from_gw([type])`

The function checks if the sender of the message is a gateway from a certain group.

This function can be used from REQUEST_ROUTE, FAILURE_ROUTE and ONREPLY_ROUTE

The function can take one optional parameter:

type (optional) - GW/destination type to be checked
flags - if message is a request and the GW has a STRIP defined, then apply it if GW is source.

Example 1.18. is_from_gw usage

...
if (is_from_gw("1") {
}
...

4.5. `is_from_gw( type, [flag])`

The function checks if the sender of the message is a gateway from a certain group.

This function can be used from REQUEST_ROUTE and FAILURE_ROUTE.

The function can take two parameters:

type (mandatory) - GW/destination type to be checked
flags (optional) - if message is a request and the GW has a STRIP defined, then apply it if GW is source.

Example 1.19. is_from_gw usage

...
if (is_from_gw("3","1") {
}
...

5. Exported MI Functions

5.1. `dr_reload`

Command to reload routing rules from database.

It takes no parameter.

MI FIFO Command Format:

		:dr_reload:fifo_reply
		_empty_line_

6. Installation

The module requires 3 table in Kamailio database: dr_groups, dr_gateways, dr_rules. The SQL syntax to create them can be found in drouting-create.sql script in kamctl db directories. You can also find the complete database documentation on the project webpage, http://www.kamailio.org/docs/db-tables/kamailio-db-devel.html.

Chapter 2. Developer Guide

The module provides no function to be used by other Kamailio modules.

Dynamic Routing Module

Edited by

Anca-Maria Vamanu

Chapter 1. Admin Guide

1. Overview

1.1. Introduction

1.2. Features

1.3. Performance

1.4. Routing Rule Definition

1.4.1. Gateway Addresses

1.4.2. Destination/GW lists

1.4.3. Routing Rules

1.5. Routing Rule Processing

2. Dependencies

2.1. Kamailio Modules

2.2. External Libraries or Applications

3. Exported Parameters

3.1. db_url(str)

3.2. drd_table(str)

3.3. drr_table(str)

3.4. drg_table(str)

3.5. drl_table(str)

3.6. sort_order (int)

3.7. ruri_avp (str)

3.8. attrs_avp (str)

3.9. use_domain (int)

3.10. drg_user_col (str)

3.11. drg_domain_col (str)

3.12. drg_grpid_col (str)

3.13. fetch_rows (int)

3.14. force_dns (int)

4. Exported Functions

4.1. do_routing("[groupID]")

4.2. use_next_gw()/next_routing()

4.3. goes_to_gw([type])

4.4. is_from_gw([type])

4.5. is_from_gw( type, [flag])

5. Exported MI Functions

5.1. dr_reload

6. Installation

Chapter 2. Developer Guide

3.1. `db_url`(str)

3.2. `drd_table`(str)

3.3. `drr_table`(str)

3.4. `drg_table`(str)

3.5. `drl_table`(str)

3.6. `sort_order` (int)

3.7. `ruri_avp` (str)

3.8. `attrs_avp` (str)

3.9. `use_domain` (int)

3.10. `drg_user_col` (str)

3.11. `drg_domain_col` (str)

3.12. `drg_grpid_col` (str)

3.13. `fetch_rows` (int)

3.14. `force_dns` (int)

4.1. `do_routing("[groupID]")`

4.2. `use_next_gw()/next_routing()`

4.3. `goes_to_gw([type])`

4.4. `is_from_gw([type])`

4.5. `is_from_gw( type, [flag])`

5.1. `dr_reload`