In some cases it may be required or useful if AVR Connection Center performs some operations differently than usual. Features on page Performance offer several such possibilities.

 

 

Maximum number of records in memory determines the amunt of data loaded at once. Processing and inserting records starts once their number reaches this value and loading continues - on a separate thread - as well. This setting has a direct impact on the amount of memory required by the software.

 

Maximum number of log messages stored has also an effect on memory usage. Every plugin instance (i.e. every configuration being executed) keeps a record of this many lines of log and makes room for new entries by deleting the oldest ones if this limit is reached. This is usually only significant if a job runs out-of-turn as a result of clicking Execute now as the scheduler regularly reads and clears messages of instances it executes.

 

Some ODBC data sources (like an SQL Server but unlike an XLS file) allow the insertion of multiple records with a single transaction. The value Number of records to be inserted at once is the number of rows to be added to a target table in a batch. Keep in mind that while inserting several records together usually improves performance we may get more comprehensible error messages when editing and testing configurations if this is set to 1.

 

Retry count shows how many attempts the software shall make at reissuing insert commands that have failed. This may prove particularly useful to eliminate random errors caused by timeouts or the target database being temporarily unreachable.

 

Enabling Stop data transfer on error halts the process of transferring data immediately if an error occurs.

 

Check Do not re-run this configuration until every other configuration in the same group is done to ensure that the scheduler only executes given configuration again if all other configurations in the same group have run since the last time it was executed (except for those that neither have nor will run because of their prerequisites).

 

Clicking Load defaults fills input boxes of the page with default numeric values.

 

Use feature Remember processed records to avoid the need to loop through rows of the source table that have already been processed at every execution. This can save much time when dealing with large amounts of data. You need to specify a column holding incremental (but not necessarily unique) values so if another record is added to the table it will have a value in this field that is greater than those of older records. Automatically incrementing IDs (AUTO_INCREMENT, IDENTITY) or columns containing the timestamps when rows have been added usually qualify as incremental.

This setting causes the plugin to record the value from the selected column of the last processed source row and will only load rows having a value equal to or greater than this in their corresponding column the next time it executes. This also means that adding new rows to the source table with a less value in this field (denoting the column isn't actually incremental) leads to the plugin never processing it as it believes it already has.

The software suspends remembering records if it encounters an error during an insert operation so that on next execution it can continue working with the row that could not have been processed because of the error. This may make it reasonable to use a Retry count greater than zero or turn Stop data transfer on error on so the plugin won't spend time processing further rows as they will have to be iterated through on next run anyway beacuse they wouldn't be remembered.

Clicking Reset feature clears information on the last remembered row so next time the entire source table will be loaded again. Selecting another incremental column leads to the same thing since the remembered value cannot be compared to those of another column.