dariuszseweryn commented on PR #9737: URL: https://github.com/apache/nifi/pull/9737#issuecomment-2675213568
> That actually highlights a larger concern in that having the Processor hold on to the Session Factory in this way is generally poor practice and should be revisited. Given the Scheduler produces records at arbitrary moments we can: 1. On each `onTrigger()` substitute the factory in the processor regularly 2. Buffer the records and submit them only `onTrigger()` — this would need to have manual checkpointing, not sure yet how cumbersome will it be > Starting and stopping is less frequent, but the framework calls onTrigger() very frequently, thus inefficient nature of the synchronized block in a method that the framework calls repeatedly. The synchronization is only used if there is no `Scheduler` created yet, it uses (searches notes) _double-checked locking_ to avoid the synchronization overhead -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
