ScadaLTS supports four types of point value logging

When point value changes is the default logging setting. The point value and its time of occurrence is written to the database only if the value of the point actually changes. This setting provides the best compromise of quality of historical information vs storage space efficiency. For Numeric points, a Tolerance, or "deadband", can also be provided; the value will be logged if the absolute value of the difference between the present value and the previous value equals or exceeds the given tolerance.

The All data setting causes every point update to be saved to the database. Only use this setting if knowledge of every update is definitely required since a great deal of data can be accumulated very quickly, which over time can cause serious performance degradation.

The Do not log setting prevents any historical data for the point from being stored in the database. Please note that chart renderers typically require logged data to function properly. Only use "Do not log" when certain that the value history is not needed.

The Interval setting allows the collection of data via the data source to be separated from its logging. Use this when the data source is collecting data very frequently (say, for event detector purposes), but all the data is not required to be logged. The Value type may be chosen with this option. "Instant" means that the current value of the point is logged, while the other options will cause a statistic of the points values over the period to be logged.

The When point timestamp changes setting. This is similar in behaviour to the on value change setting, but the timestamp of the sample is compared instead of the value.

Value discarding can be set to deal with equipment that can occasionally send errant values. When Discard extreme values is set, the Discard low limit and Discard high limit text boxes are enabled. Values less than the low limit or higher than the high limit are thereafter discarded. Any existing values already stored will remain.

The Purge strategy option allows user to choose strategy of data purging. Two types are supported: Period and Values Limit. The purge process runs automatically every morning at 3:05am local time.

The Period settings determine how long a point's historical values should be kept in the database. Values can be practically kept as long as necessary, but note that performance degradation may occur if the history for a given point becomes too large. (Actually performance of the system depends strongly upon the host hardware)

When the Values Limit type is selected, every time purging is performed for datapoint, number of values set in Values Limit will be left in database for this datapoint. The default value can be set in system settings (Values limit for purge).

The Default cache size is the number of values that will be stored in memory. This setting is especially important when a logging method other than "All data" is used. Values that are stored in memory are collated with those stored to the database when necessary. Setting this value very high may improve system performance (since it may prevent a database query), but caution should be used since memory is consumed by the cache, and out of memory conditions may occur. The "Reset cache" is normally not necessary to use, but may be convenient in cases where manual deletes of point values have been done (say, using the SQL functionality).