Comtech EF Data VMS v3.12.x Vipersat - Manual (Page 413)

Appendix C - Redundancy C-3
MN/22156, rev 12 VMS Redundancy
Redundant Hot-Standby
In a redundant configuration, the VMS servers run in parallel. The VMS data-
base on the standby server(s) is continuously maintained, in real-time, as a
mirror image of the VMS database running on the active server.
Note: It is recommended that all servers be co-located at the same site and be
connected to the same Ethernet LAN. The monitoring workstation should
also be co-located. This is to eliminate reliability issues that may be
associated with the terrestrial data-link communications between a
geographically remote server and NOC units. A data-link failure may
result in contention of automatic switch-over control and interruption of
restoral processing.
Protection Switch-over
If the active server fails, the VMS protected by N:1 redundancy immediately
switches to a standby server. The VMS running on the standby server picks up
and executes the ongoing network management tasks until the failure in the
active VMS server is resolved by human intervention.
Both the active and standby servers operate in a query-peer mode to determine
which server is to be the active VMS server in the network.
If, for example, the active VMS server fails causing a protection switch, a
standby VMS server assumes control of the network. While the standby server
is actively managing the live network, a previously active server that is being
restarted cannot assume the active server role without first checking for the
presence of an active VMS server already managing the network. The process
for initiating and managing the transitions between active to standby modes is
described below.
Active to Standby Switch
This transition occurs whenever:
• An automatic switch-over is triggered by the failure detection mechanism
due to active VMS failure, or
• A manual switch-over is invoked from the active console by, for example,
taking down the active server for maintenance.
A switch-over from the currently active server back to the server with higher
priority (once recovered) is NOT automatic. An operator must manually
perform the switch at the active server’s console.
When a server with a higher priority is restarted, the VMS on the server detects
an active peer on the network (a previous standby server) and automatically
enters standby mode, and remains in standby mode until either an operator
NOTE