You can create a group of daemons within Ecosystem Manager to monitor their group state. When you monitor the availability of daemons, you can make better decisions about resource allocation.
Each daemon group consists of a daemon that occurs in multiple servers. All daemons in the daemon group must match the executable of the daemon. A daemon cannot be part of more than one daemon group.
Active | The daemon is configured to run in an Active state and is running. |
Standby | The daemon is configured to run in a Standby state and is running. (In some cases, the daemon is configured to Standby, but is not running.) |
Passive | The daemon is not configured to run and it is not running. |
Out of Service (OOS) | The daemon has stopped running or it is dysfunctional (running and expected to be in an Active or Standby state). In most cases if the daemon is OOS, there is an issue. |
A daemon group is assigned an HA pattern which is either Active/Active or Active/Standby. If you want more than one daemon to be active on the associated servers for activities you would choose Active/Active. If you want only one server to be available for activities, with a fully redundant server kept up to date in a Passive/Standby state until needed, you would choose Active/Standby. For example, the Viewpoint postgres database is always assigned the Active/Standby pattern. The activity is updated on the single (Active) database with a complete copy maintained on the passive (Standby) database which becomes Active when the current Active database is no longer available.
Each HA pattern requires state transitions that you select and must be implemented in the state change control script. You can select other state change transitions that you set up in the state change control script. Set up the state change control script before adding daemon groups.
You can also create alerts to monitor CPU, memory usage, and heartbeats for each daemon.