17.2 Comparison of approaches to NNMi data resilience
The following table compares several aspects of these two approaches to NNMi data resilience.
Comparison item |
NNMi application failover |
NNMi running in an HA cluster |
---|---|---|
Required software products |
NNMi |
|
Time to failover |
Time needed to process the transaction logs (under normal conditions, 10 to 60 minutes) |
Under normal conditions, 5 to 30 minutes |
Transparency of failover |
Partial. The IP address of the NNMi management server changes to the physical address of what was the standby server. Users must use the new IP address to connect to the NNMi console. |
Complete. All connections use the virtual IP address of the HA cluster, which does not change in the event of failover. |
Relative proximity of active and standby servers |
LAN or WAN |
LAN or WAN (some HA products only) |
Interaction with global network management |
|
|
NNMi maintenance |
NNMi must be taken out of the application failover cluster before applying a patch or upgrading. |
NNMi can be patched and upgraded without unconfiguring HA. |