- each import flow should have a dedicated ElasticsearchImportJobExecutor instance (own queue and executor threads)
Currently one global ElasticsearchImportJobExecutor is shared between all import flows (completed process instance, running processInstance ...). This limits the throughput of the import to elasticsearch in the way that expensive write imports (user task, activityInstance & variables) tend to fill up the queue.
This is the controller panel for Smart Panels app
- mentioned in