Go to primary content
Oracle® Communications Diameter Signaling Router Policy and Charging Application
Release 8.4
F12292
Go To Table Of Contents
Contents

Previous
Previous
Next
Next

MP Server Redundancy

Redundancy models are supported for MP servers, whether deployed as DA-MPs or SBR MPs:
  • DA-MP Multi-Active Cluster

    PCA MP servers are deployed using an Active/Active redundancy model. This means that every DA-MP actively processes Diameter signaling. In order to avoid single points of failure, a minimum of two DA-MPs must be deployed (except for lab and trial deployments, where one DA-MP is acceptable). DA-MPs at a given site must be configured such that loss of a single DA-MP will not cause the remaining DA-MP servers to go into signaling overload.

  • SBR Active Only

    An SBR (either Session or Binding) can be deployed in simplex redundancy mode only for labs or trials. Otherwise this configuration represents a single point of failure for the SBR database being hosted by the Active-only Server Group. In this configuration, the SBR Server Groups consist of a single Server.

  • SBR Active/Standby

    The Active/Standby redundancy model should be used for single site PCA deployments, or for multi-site deployments when site redundancy is not important. In this configuration, the SBR Server Groups consist of two servers. On system initialization, one of the two servers in each SBR Server Group will be assigned the Active role and the other the Standby role. These roles will not change unless a failure or maintenance action causes a switch-over. For Active/Standby Server Groups, switch-overs are non-revertive, meaning that recovery of a formerly Active server will not cause a second switch-over to revert the Active role to that server.

  • SBR Active/Spare

    The Active/Spare redundancy model can be used for mated pair deployments in which it is acceptable for traffic to move from one site to the mate site on failure of a single server. In this configuration, the SBR Server Groups consist of two servers with one marked as Preferred Spare. On system initialization, the server not marked as Preferred Spare will be assigned the Active role and the other the Spare role. These roles will not change unless a failure or maintenance action causes a switch-over. For Active/Spare Server Groups, switch-overs are revertive, meaning that recovery of a formerly Active server will cause a second switch-over to revert the Active role to that server.

  • SBR Active/Standby/Spare

    The Active/Standby/Spare redundancy model should be used for PCA mated pair deployments in which site redundancy is desired. In this configuration, each SBR Server Group is configured with two servers at one site and the third at the mate site. The server at the mate site is designated in the Server Group configuration as Preferred Spare. On system initialization, one of the two servers that are located at the same site will be assigned the Active role and the other the Standby role. The server at the mate site will be assigned the Spare role (as was preferred). If the Active server can no longer perform its function due to failure or maintenance, the Standby Server will be promoted to Active. Only if both Active and Standby servers at a site are unable to perform their function will the Spare server at the mate site be promoted to Active. Active and Standby role changes within a site are non-revertive, but if the server at the mate site is Active and one of the other servers recovers, a switch-over will occur to revert the Active role back to the site with two servers.

  • SBR Active/Standby/Spare/Spare

    The Active/Standby/Spare/Spare redundancy model should be used for PCA mated triplet deployments. In this configuration, each SBR server group is configured with two Servers at one site and one server at each of two mate sites. The Server at each mate site is designated in the Server Group configuration as Preferred Spare. on system initialization, one of the two Servers that are located at the same site will be assigned the spare role (as was preferred). If the active server can no longer perform its function due to failure or maintenance, the standby server will be promoted to active. Only if both active and standby servers at a site are unable to perform their function will a spare server at a mate site be promoted to active. Active and standby role changes within a site are non-revertive, but the server at a mate site is active and one of the servers at the site with two servers recovers, a switch-over will occur to revert the active role back to the site with two servers.