Migration example: collapsing three CS 1000s into one cluster

Last Updated : Nov 16, 2021 |
Prolog information
Metadata
Keywords:
The following description is an example of the migration of a three-site CS 1000 installation into a geographically redundant Device Adapter cluster with branch survivability.

Existing CS 1000 installation

The existing CS 1000 installation consists of three sites: Main Site, M Site, and Q Site.
  • The Main Site migrates from CS 1000.1
  • The M Site migrates from CS 1000.2
  • The Q Site migrates from CS 1000.3

Migration

The three sites of the existing CS 1000 installation have duplicated TN ranges. However, not all sites have the same duplication.
The following TNs are used in this example:
  • Main Site – uses 100 0 x x (including 100 1 x x), 108 0 x x, and 116 0 x x
  • M Site – uses 104 0 x x and 124 0 x x
  • Q Site – uses 100 0 x x, 104 0 x x, and 108 0 x x
This means Q Site has three duplicated TN ranges: 100 0 and 108 0 with the Main Site 104 0 with M Site and because Q Site has all the duplications, it has been decided to change the TNs on Q Site only.
Mapping duplicate TN ranges to new ranges is implemented as follows:
TN range
Main Site
M Site
Q Site
100 0 x x
Left as
100 0 x x
Not used
Moved to
112 0 x x
104 0 x x
Not used
Left as
104 0 x x
Moved to
120 0 x x
108 0 x x
Left as
108 0  x x
Not used
Moved to
128 0 x x
112 0 x x –
Not initially used
Not used
Not used
Used for Q Site
116 0 x x
Only used here.
Left as 116 0 x x
Not used
Not used
120 0 x x –
Not initially used
Not used
Not used
Used for Q Site
124 0 x x
Not used
Only used here.
Left as 124 0 x x
Not used
128 0 x x –
Not initially used
Not used
Not used
Used for Q Site
The numbers of the three sites analyzed do not have any duplications. Furthermore, the dial plan in all three sites provides direct dialing using the extensions of the site. That is, if 555xxxx is on the Main Site, 556xxxx is on the M Site, and 557xxxx is on the Q Site, the Main Site user 5551234 can dial 5561234 to reach the M Site user. Creating a dial plan is based on this model.

Post-migration topology

The sites are configured as a single cluster in geographically redundant data centers. The data centers have their own cluster, but also have a failover cluster for the other data center.
For example, data center 1 is the data center for the three sites that aligns for the Sunny Day scenario.
  • All resources in data centers are engineered to be able to handle all endpoints from the solution across all sites.
  • A System Manager is configured in each data center as a geographically redundant pair.
  • A Duplex Communication Manager is configured per data center.
  • A pair of Session Manager is configured in each data center.
  • A high availability Device Adapter cluster is configured in each data center.
    • Two Profile 4 Virtual Machines provide highly available Device Adapter service for up to 5000 endpoints.
    • Each additional Profile 4 Virtual Machine adds support for an additional 5000 endpoints.
  • Additional resiliency can be configured for improved resiliency for MGCs:
    • An LSP / Survivable Core with a local (scaled down) Device Adapter is configured.
    • A Branch Session Manager is configured.
    • A high availability Device Adapter cluster is configured, with a suitable size.
    • All resources in the survivable site are engineered to be able to handle only MGCs that fail over to this data center.
Note:
This migration involves three CS 1000 systems. Using one Device Adapter cluster with geographical redundancy minimizes the migration administration.
The following diagram is the topology of the post-migration configuration.