1 Introduction
This section is an introduction to this document. It contains information about the prerequisite, purpose, scope, and target group for the document. This section also contains explanations of typographic conventions used in this document.
1.1 Purpose and Scope
This document describes the Administration of Multi Regions and BSS Capacity feature in the Ericsson™ Dynamic Activation (EDA).
1.2 Target Group
The target group for this document is as follows:
- Network Administrator
- Application Administrator
- Other
For information about the different target groups, see Library Overview, Reference [1]
1.3 Typographic Conventions
Typographic conventions are described in Library Overview, Reference [1].
2 General
Administration of Multi Regions and Administration of Business Support System (BSS) Capacity are two features in the Dynamic Activation Provisioning Manager functional group. These two features have technical dependency with other features in Dynamic Activation and in the layered HLR.
As illustrated in Figure 1, the GSM/UMTS feature in Dynamic Activation is one of the three mandatory components in the UDC solution for the HLR application. Logical HLR and Multi Region are two HLR optional features. These features are aiming for segmenting the HLR capacity in the UDC solution for the operator but in different flavors. With either of them, operators can introduce the optional feature Administration of Multi Regions in the network. Therefore operators can manage subscriber data in a segmented manner. On top of this, another optional feature Administration of BSS Capacity can be deployed, providing the mechanism to regulate provisioning capacity use between the provisioning clients. The Administration of Multi Regions feature is the prerequisite for the Administration of BSS Capacity feature.
For more information about the provisioning solution for HLR in data layered architecture, see Function Specification Layered HLR, Reference [2]. For more information about the software platform Provisioning Server that hosts these two features, see Function Specification Resource Activation, Reference [3]. For information about configuring these features, see Configuration Manual for Resource Activation, Reference [4].
3 Administration of Multi Regions
This feature enables sharing of UDC HLR network infrastructure among different regions/countries/service providers. The key functions are:
- Management of Administration Domains (then called Admin Domains): This is the foundation for the next two functions. Defined Admin Domains must reflect the service agreement between the network infrastructure owner and the region/countries/service providers.
- Access control: This function ensures that each region/country/service provider only can administer its own subscribers; other subscribers served by the same network infrastructure are not accessible. Every provisioning request is evaluated towards the restriction rules defined in the Admin Domain. Provisioning requests that violate the restriction rules are rejected. This also ensures that a subscriber only can access the services provided within its region/countries/service providers, regardless of the network capability.
- Manage Region Identifier (RID) for BSS during provisioning: The attribute RID can be handled transparently by Dynamic Activation. Thus the BSS is not affected.
3.1 Management of Administration Domains
An Admin Domain is a special provisioning role with mandatory restriction rules. An Admin Domain is defined with the purpose to group subscribers within a network. It is possible to:
- Add/Edit/View an Admin domain
- Compare two Admin Domains
- Associate an Admin Domain to a user. For information about user authorization, see Function Specification Resource Activation, Reference [3].
When defining an Admin Domain, the following rules must be considered:
- Up to 32 Admin Domains are supported.
- Each Admin Domain can be associated with one or more users.
- Each Provisioning Client can be associated with one
or more Admin Domains. The following is valid within the Admin Domain
context:
- A user having access to only one Admin Domain is called a Provisioning Client.
- A user having access to more than one Admin Domain is called a Local Administrator.
- A user having access to all Admin Domains and the System Administrator Admin Domain is called a System Administrator.
- All Admin Domains must have restriction rules for the mandatory key parameters - RID and IMSI.
- MSISDN number series restriction rule is a special rule.
- Admin Domains can never have overlapping IMSI number series, MSISDN number series or RID values. Thus a subscriber belongs to only one Admin Domain.
- In addition to the key parameters, Admin Domains can contain restriction rules for other attributes, for example, Subscriber Data Profile IDs attribute.
3.2 Use of Administration Domains in Provisioning
The following two chapters describe how Admin Domain and RID are used in Individual provisioning requests and Massive provisioning requests. CAS backwards compatibility is achieved by not including RID.
3.2.1 Individual Provisioning
Individual provisioning request scenarios are outlined in Figure 2.
BSS can choose to include or exclude the RID in the provisioning requests. When RID is not included in the provisioning requests, the logic locates it and maps it to the request:
- The provisioning requests contain IMSI: As IMSI is one of the mandatory attributes in Admin Domains, it points to one Admin Domain and thus the RID.
- The provisioning requests contain only the MSISDN as identity: The logic first uses the MSISDN to locate the Admin Domain. If the MSISDN is not found (because of the presence of MNP), a CUDB lookup based on MSISDN is initiated to find the IMSI. The returned IMSI is then used to locate the Admin Domain and then the RID.
During the access control process:
- If both user and Admin Domains restriction rules exist, the request is evaluated towards the user restriction rules, and then the Admin Domain ones.
- An attribute validated against the User authority is not validated against the Admin Domain authority. In other words, user restriction rules override the Admin Domain restriction rules.
- If a user is associated to more than one Admin Domain, access control validates the request against the authority of all the Admin Domains. The request must match the authority of one of the Admin Domains.
- A user can only administer the subscribers defined in the Admin Domain to which it belongs.
- Requests are rejected if the RID/MSISDN/IMSI combination in the request conflicts with that defined in the Admin Domain.
This is an example of how the Admin Domain can be used. Table 1 defines a setup for a small operator with restriction rules and its associated users.
|
Name |
Restriction Rules |
Capacity Attributes and other Attributes |
Users |
|
Administration Domain 1 |
RID=1 |
user_one_prov, user_11_IN, user_21_OSS, user_localadm, user_sysadm | |
|
Administration Domain 2 |
RID=2 |
user_two_prov, user_12_OSS, user_localadm, user_sysadm | |
|
Administration Domain 3 |
RID=3 |
user_three_prov, user_13_OSS, user_localadm, user_sysadm | |
|
Administration Domain 4 |
RID=4 |
user_four, user_14, user_sysadm | |
|
Administration Domain System Administrator |
sysadm |
user_sysadm |
Examples of Individual provisioning request behavior:
Example 1. Create subscriber without RID
log in as: user_one_prov
> HGSUI:IMSI=123456789012345,MSISDN=345678901234567;
<OK - Subscriber created>
The request is associated with Admin Domain 1 because of matching IMSI. The command passes Admin Domain ACL because user_one_prov is associated with Admin Domain 1.
Example 2. Create subscriber without RID and wrong user
log in as: user_12_OSS
> HGSUI:IMSI=123456789012345,MSISDN=345678901234567;
<Error message - user has no access to Admin Domain 1>
The request is associated with Admin Domain 1 because of matching IMSI. The command fails in Admin Domain ACL because user_12_OSS is associated with Admin Domain 2.
Example 3. Create subscriber with RID
log in as: user_one_prov
> HGSUI:IMSI=123456789012345,MSISDN=345678901234567,RID=2;
<Error message - RID mismatch>
The request is associated with Admin Domain 1 because of matching IMSI. The command fails in Admin Domain ACL because RID=2 mismatch with RID=1 from Admin Domain 1.
Example 4. Print Subscriber Data with MSISDN
Subscriber with MSISDN=345678901234567 and IMSI=123456789012345 exists in CUDB.
log in as: user_one_prov
> HGSDP:MSISDN=345678901234567;
<OK - prints data of subscriber>
As MNP is presented in the network, MSISDNS is not configured in the Admin Domains. A CUDB lookup is performed to get the IMSI. The request is associated with Admin Domain 1 based on the retrieved IMSI. The command passes Admin Domain ACL because user_one_prov is associated with Admin Domain 1.
3.2.2 Massive Provisioning
Massive provisioning Admin Domain association and Access Control flow are described in figure Figure 3. There are only two variants of Massive provisioning requests - with RID and without RID.
Massive provisioning requests have the following characteristics:
- A Massive request is associated with one or several Admin Domains based on the logged-in user or RID provided in the request, not the IMSI, or MSISDN keys as in Individual provisioning. If RID is omitted, the request is associated with all Admin Domains the logged-in user is associated with, see Example 1 exMass01 and Example 4 on exMass04.
- If RID values and key values, for example MSISDN or IMSI, mismatch in the Admin Domain ACL, the requests is not blocked. The purpose is to maintain a strong search flexibility regarding combinations of RID and key values, for example IMSIS and MSISDNS. However, if the RID value mismatch with the logged-in user values, the Admin Domain ACL blocks the request, see Example 2, 3, and 5 on exMass02.
- A user can be associated to more than one Admin Domain. The user can:
There are several examples of Massive provisioning requests, which clarify Admin Domain association and Access Control behavior.
Admin Domain configurations in these examples are the same as in the Individual provisioning examples, see Table 1.
Example 1. Massive search for all subscribers that match IMSIS=12345688
log in as: user_one_prov
> HEMSSDP:IMSIS=12345688,SSDA;
<OK, Returns a number of users matching IMSIS=12345688 and RID=1>
> HEMSSDP:IMSIS=12345688,SSDA,RID=1;
<OK, Returns a number of users matching IMSIS=12345688 and RID=1>
Both the requests print subscribers from Admin Domain 1 with IMSIS=12345688. The first request is assigned to RID=1 during processing because user_one_prov is associated with Admin Domain 1. The second request has an explicit RID=1.
Example 2. Print all subscribers that match IMSIS=222222
log in as: user_one_prov
> HEMSSDP:IMSIS=222222,SSDA;
<OK, Does not return any users>
> HEMSSDP:IMSIS=222222,SSDA,RID=1;
<OK, Does not return any users>
Both requests do not print any users because the CUDB combined search criteria IMSI and RID do not match any users.
Example 3. Print all subscribers that match IMSIS=123456 and RID=2 (Admin Domain 2)
log in as: user_one_prov
> HEMSSDP:IMSIS=123456,SSDA,RID=2;
<Error message - user has no access to Admin Domain 2>
Returns an error because of RID=2 and associated Admin Domain 2 is not associated with logged in user user_one_prov.
Example 4. Change all subscribers that belong to Admin Domain 1
log in as: user_one_prov
> AEMSSUC:IMSIALL,AMF=14532,FSETIND=2;
<OK, Change subscribers that belong to Admin Domain 1>
> AEMSSUC:IMSIALL,AMF=14532,FSETIND=2,RID=1;
<OK, Change subscribers that belong to Admin Domain 1>
Both requests change all subscribers that are associated with Admin Domain 1 which is associated with user_one_prov.
Example 5. Change all subscriber that belongs to Admin Domain 1 and 2
log in as: user_one_prov
> AEMSSUC:IMSIALL,AMF=14532,FSETIND=2,RID=1&2;
<Error message - user has no access to Admin Domain 2>
Requests return an error because user_one_prov have no access to Admin Domain 2.
The following examples show how System Administrator and Local Administrator can use RID and number series to define search scope to cover more than one Admin Domain. Admin Domain configurations in these examples are the same as in the Individual provisioning examples, see Table 1.
Example 1. Print all subscribers that user_localadm has access to
log in as: user_localadm
> HEMSSDP:IMSIS=1234,SSDA;
<OK, Output>
Prints all subscribers user_localadm has access to which are Admin Domain 1, 2 and 3. RID=1&2&3 is associated with the request because user_localadm is associated with Admin Domain 1, 2 and 3 and thus limits the number of returned subscribers. The IMSIS=1234 would include all subscribers in the network defined in Table 1, but the result is limited which is based on what Admin Domain associations user_localadm has.
Example 2. Print all subscribers from Admin Domain 1 and 3
log in as: user_localadm
> HEMSSDP:IMSIS=1234,SSDA,RID=1&3;
<OK, Prints subscribers from Admin Domain 1 and 3>
Prints all subscribers from Admin Domain 1 and 3.
Example 3. Change all subscribers from Admin Domain 1 and 3 using IMSIALL instead of IMSIS
log in as: user_localadm
> AEMSSUC:IMSIALL,AMF=14532,RID=1&3;
<OK, Changed subscribers from Admin Domain 1 and 3>
Example 4. Change all subscribers user_localadm has access to
log in as: user_localadm
> AEMSSUC:IMSIALL,AMF=14532;
<OK, Changed subscribers from Admin Domain 1, 2 and 3>
Changes all subscribers user_localadm has access to, which are from Admin Domain 1, 2 and 3.
Example 5. Change users with IMSIS=1234 and RID=2&4
log in as: user_sysadm
> HEMSSDP:IMSIS=1234,SSDA,RID=2&4;
<OK, Prints all subscribers from Admin Domain 2 and 4>
Prints all subscribers from Admin Domain 2 and 4. RID is provided, that limits the resulting scope of subscribers.
Example 6. Print all subscribers with IMSIS=1234, whole network
log in as: user_sysadm
> HEMSSDP:IMSIS=1234,SSDA;
<OK, Prints subscribers from the whole network>
Prints all subscribers in the whole network. The IMSIS=1234 include all subscribers in the network defined in Table 1.
4 Administration of BSS Capacity
This feature ensures fair provisioning capacity sharing between regions/countries/service providers, thus avoiding Denial of Service.
Fair capacity sharing is enforced by introducing Capacity Attributes that define restrictions on the provisioning resources which a provisioning client or provisioning role can utilize. Two Capacity Attributes are used in Dynamic Activation:
- Connection Restriction Control: Defines the maximum number of allowed parallel logons on the provisioning interfaces. When the configured limit on a given interface is exceeded, new logon attempts are rejected with an error message.
- HLR Sustainable Provisioning Throughput Restriction
Control: Defines the minimum HLR Sustainable Customer Service
Orders per second. Can also be described as Maximum Allowed HLR Customer
Service Orders per second under saturated traffic conditions. The
enforcement mechanism is only enabled during saturated traffic conditions.
Feature administration of BSS Capacity depends on feature administration of Multi Regions.
4.1 Connection Restriction Control
The feature enables the possibility to configure the Capacity Attribute "maximum number of allowed parallel logons" on the MML and CAI3G interfaces. The restrictions apply per user and Dynamic Activation system.
One limit can be configured for MML (includes both TELNET and SSH), and CAI3G (includes both HTTP and HTTPS). They are mutually independent. One user can be allowed to have, for example five parallel logons on the MML interface, and 10 parallel logons on the CAI3G interface simultaneously.
When the configured limit on a given interface is reached, a new logon attempt is rejected with an error message.
The configuration and restriction check is system wide, which means that all logons on different cluster nodes add up towards the configured maximum limit.
For example, if user_one_prov has maximum five MML sessions, the sixth time logon attempt to the Dynamic Activation over the MML interface is rejected.
When defining connection restrictions, all provisioning clients must be considered. This means that the connection capacity of a system must be shared by users not only provisioning the HLR, but also other applications, such as IMS.
The MML and CAI3G protocol implementations are synchronous. This means that every MML client and CAI3G client have only one outstanding request per client. The following table illustrates the theoretical restriction figures that can be expected using the Connection Restriction feature.
|
User |
Number of MML Connections/Number of CAI3G HTTP Clients |
Example of Latency in Milliseconds |
Throughput per User and MML Connection or CAI3G HTTP Client (CSO/sec) |
|---|---|---|---|
|
user_one_prov |
10 |
100 |
100 |
|
user_11_IN |
5 |
100 |
50 |
|
user_localadm |
1 |
100 |
10 |
|
user_two_prov |
20 |
100 |
200 |
4.2 Sustainable Provisioning Throughput Restriction Control
The purpose of the HLR sustainable throughput Restriction Control is to limit throughput on Individual Provisioning interfaces - MML and CAI3G. The feature has the following characteristics:
- The throughput restriction is set and measured in Customer Service Orders per second - CSO/sec.
- Restrictions can be set on both Provisioning Clients and Admin Domains, see Section 3.
- The configuration and throughput restriction check is cluster wide, which means that all requests on different payload nodes add up towards the configured maximum throughput limit.
- Each system has a maximum traffic threshold which depends on the number of payload nodes in the system. The restriction is enabled only when maximum traffic threshold is reached. This means Provisioning Clients or Admin Domains could use more throughput than the configured values if the threshold is not reached.
- The total configured CSO/sec value for all Provisioning Clients belonging to one Admin Domain must never exceed the Administration Domain configured CSO/sec.
- The subscriber identifier MSISDN or IMSI defines which Admin Domain a request belongs to. Common Data or Massive Service DATA provisioning is not restricted.
- If there are other applications that provision through Dynamic Activation, such traffic is measured when calculating the maximum traffic threshold. However, when the threshold is reached, throughput control applies only on HLR and AUC individual provisioning requests.
- The Sustainable Provisioning Throughput Restriction Control is a best effort implementation. Configured throughput values and observed throughout values for a single user or domain can differ significantly when more than three users or domains are configured. Combine this feature with the Connection Restriction Control feature to fine-tune the throughput.
There are several examples that illustrate how previously characteristics behave.
Example 1. Throughput is lower than traffic threshold.
|
Number of Payload Nodes |
Traffic Threshold CSO/sec |
Admin Domain |
Min. HLR Sustainable Throughput per Domain CSO/sec |
User |
Min. HLR Sustainable Throughput per User CSO/sec |
|---|---|---|---|---|---|
|
4 nodes |
1200 |
1 |
400 |
user_one_prov |
- (not specified) |
|
2 |
800 |
user_two_prov |
- |
- user_one_prov sends 450 CSO/sec.
- user_two_prov sends 500 CSO/sec.
No restrictions are enforced because system has not reached traffic saturation.
Example 2. Admin Domain throughput restrictions enforced when throughput has reached traffic threshold.
|
Number of Payload Nodes |
Traffic Threshold CSO/sec |
Admin Domain |
Min. HLR Sustainable Throughput per Domain CSO/sec |
User |
Min. HLR Sustainable Throughput per User CSO/sec |
|---|---|---|---|---|---|
|
4 nodes |
1200 |
1 |
400 |
user_one_prov |
- |
|
user_11_IN |
- | ||||
|
2 |
800 |
user_two_prov |
- |
- user_one_prov sends 300 CSO/sec, the sum of CSO/sec for user_one_prov and user_11_IN is limited to 400.
- user_two_prov sends 800 CSO/sec.
- user_11_IN sends 150 CSO/sec, the sum of CSO/sec for user_one_prov and user_11_IN is limited to 400.
Restrictions are enforced because system has reached traffic saturation.
- Note:
- The Connection Restriction Control feature is needed to get the exact Admin Domain throughput balance.
Example 3. One node stops to receive traffic.
|
Number of Payload Nodes |
Traffic Threshold CSO/sec |
Admin Domain |
Min. HLR Sustainable Throughput per Domain CSO/sec |
User |
Min. HLR Sustainable Throughput per User CSO/sec |
|---|---|---|---|---|---|
|
4 nodes |
1200 |
1 |
400 |
user_one_prov |
- |
|
user_11_IN |
- | ||||
|
2 |
800 |
user_two_prov |
- |
- user_one_prov sends 300 CSO/sec.
- user_11_IN sends 100 CSO/sec.
- user_two_prov sends 800 CSO/sec.
Node one stops receiving traffic reducing system-wide throughput to 900 CSO/sec.
Provisioning Client and Administration Domain throughput is reduced to 900 CSO/sec and user throughput decreases proportionally to the configured Admin Domain throughput restrictions.
- Note:
- The Connection Restriction Control feature is needed to get the exact Admin Domain throughput balance.
Example 4. Provisioning Client throughput restriction when throughput has reached traffic threshold.
|
Number of Payload Nodes |
Traffic Threshold CSO/sec |
Admin Domain |
Min. HLR Sustainable Throughput per Domain CSO/sec |
User |
Min. HLR Sustainable Throughput per User CSO/sec |
|---|---|---|---|---|---|
|
4 nodes |
1200 |
1 |
- |
user_one_prov |
300 |
|
user_11_IN |
100 | ||||
|
2 |
- |
user_two_prov |
800 |
- user_two_prov sends 800 CSO/sec.
- user_one_prov sends 300 CSO/sec
- user_11_IN tries to send 150 CSO/sec, but gets only 100 CSO/sec because the Admin Domain 1 CSO/sec = 400 and user restriction is set to CSO/sec = 100.
Restrictions are enforced because system has reached traffic saturation.
- Note:
- The Connection Restriction Control feature needs to get the exact Admin Domain throughput balance.
Example 5. Throughput restriction for a user belonging to more than one Admin Domain when throughput has reached traffic threshold.
|
Number of Payload Nodes |
Traffic Threshold CSO/sec |
Admin Domain |
Min. HLR Sustainable Throughput per Domain CSO/sec |
User |
Min. HLR Sustainable Throughput per User CSO/sec |
|---|---|---|---|---|---|
|
4 nodes |
1200 |
1 |
400
|
user_one_prov |
300 |
|
user_11_IN |
50 | ||||
|
1, 2 |
user_localadm |
- | |||
|
2 |
800 |
user_two_prov |
800 |
- user_two_prov sends 750 CSO/sec, 50 CSO/sec left of Admin Domain 2 throughput.
- user_one_prov sends 300 CSO/sec, 100 CSO/sec left of Admin Domain 1 throughput.
- user_localadm sends 100 CSO/sec, utilizing 50 CSO/sec from Admin Domains 1 and 50 CSO/sec from Admin Domain 2.
- user_11_IN sends 150 CSO/sec, but gets only 50 CSO/sec because the user restriction is set to CSO/sec = 50.
Restrictions are enforced because system has reached traffic saturation.
- Note:
- The Connection Restriction Control feature needs to get the exact Admin Domain and user throughput balance.
Example 6. Admin Domain and Provisioning Client combination throughput restriction when throughput has reached traffic threshold.
|
Number of Payload Nodes |
Traffic Threshold CSO/sec |
Admin Domain |
Min. HLR Sustainable Throughput per Domain CSO/sec |
User |
Min. HLR Sustainable Throughput per User CSO/sec |
|---|---|---|---|---|---|
|
4 nodes |
1200 |
1 |
400 |
user_one_prov |
- |
|
user_11_IN |
10 | ||||
|
2 |
800 |
user_two_prov |
- |
- user_two_prov sends 800 CSO/sec.
- user_one_prov sends 600 CSO/sec, gets 390 CSO/sec
- user_11_IN tries to send 20 CSO/sec, gets 10 CSO/sec
Restrictions are enforced because system has reached traffic saturation.
- Note:
- The Connection Restriction Control feature needs to get the exact Admin Domain and user throughput balance.
Reference List
| [1] Library Overview, 18/1553-CSH 109 628 Uen |
| [2] Function Specification Layered HLR, 4/155 17-CSH 109 628 Uen |
| [3] Function Specification Resource Activation, 3/155 17-CSH 109 628 Uen |
| [4] Configuration Manual for Resource Activation, 2/1543-CSH 109 628 Uen |

Contents


