We are briefly exceeding the query concurrency limits of our warehouse intermittently resulting in errors. I noticed that there is a setting called resource classes which can increase/decrease the concurrency by altering the memory usage of each query. From my understanding, I can decrease the memory usage to get a greater concurrency, or increase the memory usage to increase performance but get lower concurrency in return.
Is there a generally recommended Resource Class setting for Qlik Replicate? Guidance on this is appreciated. Thank you.
No magic setting on this one from what I can see. It's really going to depend on the amount of Tasks you have loading to the Synapse endpoint and how they're configured. This will be the access pattern on Synapse and you could look to tune Synapse from there. You can control the chunk/batch sizes of both Full Load and CDC in a Task so that would be a good place to look, but this could affect Task latency. I'd use Professional Services here if you have the option.
