High Availability (HA) is a feature that's designed to reduce the gaps in collected data in certain situations where Prognosis Servers would otherwise become unavailable.
Typical situations that could experience downtime include:
- Planned maintenance of server operating system or hardware where Prognosis Server is running.
- Patching or upgrading existing Prognosis Server installations.
- Any failure of critical Prognosis component on the primary-active server.
HA now allows these situations to be handled with minimal loss in data collection.
This is achieved by allowing two Management or Monitoring servers to be configured as a HA Pair, where their Configurations are synchronized but only one of the servers is Active and collecting data. The capability to switch the servers between Active and Passive modes is performed using the High Availability page in the Web Application - Administration tool.
This allows any maintenance to be performed on the Passive server which is not collecting data and hence the only downtime experienced in data collection is the time taken to switch the activity state of the HA Pair.
It is important to understand the Limitations When Using HA, and to understand that Database Collections (Prognosis High-Performance Database) are not synchronized using HA. The Database Collections will only be active on one server at any one time. Switching between Active and Passive modes ensures the continuity of information being provided however the data will be stored across both servers and each data collection will contain gaps where the data is stored on the other server.