2.1.19
2023-05-03
New features
Support for $batch calls on Create and Update operations
You can now select the iterator option "Batch with single change set" when creating or updating multiple entities via the iterator feature. In practice these will make the connector send all the create/update requests in one single $batch request to IFS - which has significantly better performance. The "single change set" also implies an "all or nothing"-operation - either all entities are created/updated - or none are. This is, however, dependent on the implementation of $batch on the endpoint that receives it. All tests we have done has followed this rule, but we cannot garantuee it.
Order by name in communication log. A small feature to make it easier to work with the communication log
This version of the connector is based on newer version of the connector service infrastructure which includes several new features.
Configurable http timeout. You can now configure the http timeout on poll worker services. Setting the value to lower than 100 seconds is not recommended. New installations will have the default value of 100 seconds, upgraded ones will have 43200 (12h).
Faster scaling. The connector will scale it's running workers up and down much more quickly and predictable. Each running worker consumes memory and CPU, so having the right amount of workers running is good for the overall performance of your machine hosting the connector.
Execution monitor
A new section which can be reached from here:
In this section you can monitor the part of the connector that handles communication with the external system, IFS in this case. The image below is from a load test, not normal load. The green color indicates how much requests are in the internal queue of the connector, ready to be processed. The black line shows incoming requests to the connector. This particular load test had much less number of running workers compared to incoming load, so the green area was expected to go up as shown below. Below the graphs for incoming/outgoing you can see all currently running workers, and what they are doing. You can also get sense of the execution times and how many jobs they have done.
NOTE: Do not leave the monitoring page open in production. It consumes resources from the connector.
Bug fixes
Fixed a bug where filtering on expand members caused an invalid error message
Fixed a bug where custom requests could not be used in a subscribed workflow if IFS endpoint differed. Note that when opening the machine step in subscribed workflow the url will point to the url of the published workflow. The correct url will be used at runtime though.
Fixed a bug which gave an "Filter on unknown member"-error when the name of an entity property was the same as the name of the entity
Various minor bugfixes
Last updated
