About Collector Redundancy

Collector redundancy decreases the likelihood of lost data due to software or hardware failures. It ensures that collection of your data remains uninterrupted. It uses two or more collectors that gather data from a single source.

All collectors in the group actively gather the same tags from a data source but only the active collector forwards its samples to the Historian server. The non-active (standby) collectors buffer their data against failover of the active collector. The Historian server actively monitors the health of the redundant collectors and will automatically switch to a backup if certain trigger conditions are met.

Trigger conditions: You can configure the following triggers:
  • The collector is stopped.
  • The collector sends bad quality data for a specified tag (called the watchdog tag).
  • The values of the watchdog tag do not change for a specified duration.

If any of these trigger conditions are met, the primary collector goes into the standby mode, and the secondary collector becomes active. That means, data is then collected by the secondary collector. Similarly, if the secondary collector satisfies the trigger conditions, it goes into the standby mode, and the primary collector becomes active.

You can also manually trigger the failover.

For instructions on configuring collector redundancy, along with settings the triggers, refer to Configure Collector Redundancy.

Synchronized Configuration: The configuration for each redundant collector is managed by the Historian server. If you change the configuration on one redundant collector, all other collectors will be updated automatically with the new configuration.
Note: Redundant collectors must be of the same collector type.

Offline Collectors: To reduce the possibility of lost data, a collector will immediately send its buffered data to the Historian archiver when brought online. The Historian server will ignore any data that is already collected in the archive.

Adding Tags to a Redundant Collector: To add tags in a redundant collector system, you must add them to the primary collector. If tags are added to a redundant collector, data for these tags will not be collected until the redundant collector is taken out of the redundant system. If the primary collector is not available, you must add the tag manually.
Note:

  • Use polled tags only as watchdog tags.

  • Historian redundant collector configuration does not force the active Historian collector to run on the active iFIX SCADA, since both redundant collectors provide data. Also, when both iFIX SCADAS become active, they lose connection with each other.

Important information about the failover of redundant collectors:

  • In the Redundancy section of the Collectors page, you can use the Make Active Collector Now! button to manually force a failover to a backup collector.

  • In an Enterprise system, collector redundancy failover happens only after 5 minutes after a tag change. You must select the Make Active Collector Now! button after the first 5 minutes for the failover to happen.

    Also, when you shut down an active collector, it does failover. However, if there was a tag change then shutting down the active collector does not cause failover immediately, it is just delayed by 5 minutes. There will not be any data loss since the backup collector sends data for the past 15 minutes when it becomes active.

  • Failover precedent is cyclical; the last collector in a redundant group will automatically failover to the first collector in the group.
  • Configuration Manager must be running for failover to happen in a mirrored environment.