Avaya Media Processing Server Series System (Software Release 2.1) Instrukcja Użytkownika Pobierz pdf (Strona 273)

Configuration Procedures and Considerations

# P0602477 Ver: 3.1.11 Page 273

The Failover/Failback Process

1. Upon startup in a standby configuration, each secondary TRIP attempts to

connect to all TMSs for which it is listed as secondary in the

$MPSHOME/common.standby/etc/tms/tms.cfg file. All

primary components controlling these TMSs reside on different nodes.

2. Upon receiving a signal from the TMS that it lost its primary controller

(i.e., the primary component failed), TRIP informs the local RCD that

failover is required. RCD sets its state to “failover initiated” and sets the

secondary node configuration to match the configuration of the primary

node (using backup local copies of primary MPS and COMMON

components). If the primary node has fewer components than the

secondary node, unused components are pointed to empty configuration

($MPSHOME/mps0.empty). Once failover is initiated by RCD, all

subsequent failover requests from TRIPs are ignored (i.e. the first primary

node whose failure is detected by the secondary will be taken over). After

changing configuration to the primary node’s configuration, RCD restarts

local COMMON and MPS components to force them to load the new

configuration.

3. When components are restarted, RCD waits 5 minutes for local TRIPs to

acquire TMSs. If the TRIPs cannot acquire TMSs, the TMSs may need to

be restarted. This is reported as the “grace period” in the “Local MPS

States” panel in the RCD status report. The grace period ends when

either:

•all local MPS components (excluding those loaded empty configuration)

acquired their TMSs.

•5 minutes have elapsed

4. Redundancy is implemented on a node-by-node basis. Once entering the

active state, the secondary node protects all MPS components on the

failed primary node for current and future failures but does not protect

other primary nodes.

For example a primary node runs two MPS components. When one of

them fails, secondary node will enter the active state and take over the

failed component and also will monitor the second primary component

for failures. When the second component failure is detected, the

secondary node will take it over. The RCD status report shows whether a

secondary MPS component acquired the TMS.

5. Once the failed primary node is operational again, it can request failback

(see TRIP Failback on page 268) from the secondary node. Once

failback is requested, the secondary TRIPs release their TMSs, allowing

primary TRIPs to take control. Once all TMSs are released, the secondary

node goes back to standby state, again protecting all primary nodes in the

cluster.

Alarms are displayed as systems change state. See the MPS Alarm Message Reference

Manual for information regarding the alarms and their meanings.

1 2 ... 268 269 270 271 272 273 274 275 276 277 278 ... 305 306

Komentarze do niniejszej Instrukcji

Brak uwag

Avaya Media Processing Server Series System (Software Release 2.1) Instrukcja Użytkownika Strona 273

Komentarze do niniejszej Instrukcji

Powiązane produkty i podręczniki dla Części komputerowe Avaya Media Processing Server Series System (Software Release 2.1)