Follow

What happens if I incorrectly configured my calendar?

Calendars

The foundation of Flow information is the time context against which Flow aggregates, calculates and stores data.  In order to configure how Flow creates the underlying time period context you need to create Flow Calendars and Shift Patterns.

For example, your reporting day may start at 06:00 in the morning or your reporting year may start in March.

 General Configuration

The general configuration of a calendar consists of:

  • Description:        short description of the Calendar
  • Management:        by default Flow will manage a Calendar’s time period creation automatically based on the standard Gregorian calendar system, applying the applicable offsets where required.
  • Year starts in:        used to define the month in which the reporting year will start (e.g. financial year)
  • Week starts on:        used to define the day on which the reporting week will start.
  • Day starts at:        used to define the hour and minute at which the reporting day will start.
  • Hours split into:        used to define the duration of “minutely” time periods.  It is important to note that “minutely” in Flow does not mean a single minute, but rather a sub-hourly period that has this duration.  The smallest “minutely” time period duration is 5 minutes.  If this setting is set to 60 minutes, then Flow will not allow the creation of minutely measures.

The general calendar settings can only be modified when all Flow Engines are undeployed.

Default Refresh Offset

The Refresh Offset for a measure is the “delay” or offset from the “Period End” to the time the retrieval of a measure will be performed by the Flow Engines.  For example, at the end of an hourly period, Flow will wait 60 seconds by default before querying any data sources for measure values.

Each calendar has a set of default refresh offsets that will be used when a calendar is first linked to a measure when that measure is created.  The refresh offset for that measure can be edited independently of the Calendar setting if required.

Auto Purge

Flow has the ability to automatically remove (i.e. purge) old data.  This ensures a well maintained system.  For example, if we set the minutely time period’s Purge Age to 90, the Engines will automatically remove minutely measures’ data that when that data is older than 90 days.  By default the Purge Age for each Time Interval type is set to 0 days, which implies that purging is disabled for that Time Interval.

Purging is recommended for large systems that have accumulated a few years of data.  If purging is not configured, your Flow System may perform slowly.

Purge Now

For Flow systems where the auto-purging functionality was not previously configured, and on which the performance has already been impacted, it is possible to perform an immediate purge, using the “Purge Now” functionality.

Note that purging is an irrecoverable process.  It is recommended to create a database backup before performing a Purge Now..

What if I need to Change an active Calendar?

If you need to change the start settings of your time intervals and Flow already created time buckets, you will notice that the calendar settings are grayed out. These settings can only be changed if there are no time buckets created against the calendar.

The "Purge Now" functionality can be used to delete the currently created time buckets.

The procedure to change the calendar settings are the following:

  1. Make sure to have a current backup of the Flow Database and make sure to have data that was manual captured available.
  2. Undeploy all engines and platforms
  3. Perform the "Purge Now" option by supplying the name of your Flow instance (the instance name can be located under the "System/Properties" menu item)
  4. After the purge operation, the calendar setting will be editable.
  5. After making your changes, deploy your Platform and Engines
  6. Preform a backfill operation on your measures.
  7. Please be aware that manual measure's data will be lost and will not be generated by the backfill operation.

 

Was this article helpful?
0 out of 0 found this helpful
Have more questions? Submit a request

Comments