View in ref guide.
Edit this option Publish this option Clear draft content
Published Option
backup-file-aggregates-store
Default Value: true
Valid Values: true, false
Changes Take Effect: Immediately upon notification
Dependencies: auto-backup-interval != 0
Introduced: 8.5.102
Controls whether the statistical aggregates are read from/written into the backup file.
While reading from the backup file, Stat Server might use, adjust then use, or ignore a serialized aggregate, depending on the interval type.
If the system time changes during or in between writing/reading aggregates, the aggregation window duration is adjusted approximately. Otherwise
the following cases are applicable:
| Interval |
Near past
(Aggregate is OK) |
Distant past
(Aggregate is outdated) |
|---|
| Growing | use | ignore |
| Sliding | adjust then use | ignore |
| Selection | use | use |
Notes:
- The aggregate is only used, if no important statistical attributes (TimeProfile, Filter) have changed, since it was written into a backup file.
- The aggregate serialization is not applied to Java statistics.
- If a statistic aggregates over the duration (e.g., TotalTime) and has the Subject != DNAction, then incomplete subjects are forcefully ended prior to benign written into the backup file at shutdown (or handling the HistoryLogExpired); otherwise, incomplete subjects are not written into the backup file at shutdown (or while handling the HistoryLogExpired). If a writing into the backup file happens upon a timer, specified by the auto-backup-interval option, no incomplete subjects are written for any statistics.
Draft Option
No draft option
