paul-rogers commented on issue #1455: DRILL-6724: Convert IndexOutOfBounds exception to UserException with … URL: https://github.com/apache/drill/pull/1455#issuecomment-418200233 I have a basic question on this one. Is an Index Out Of Bounds Exception (IOBE) an expected behavior for this context? Or, does it indicate a bug in the code? Normally, code is expected to know the array size, and index only within that size, meaning that an IOBE is a program error. By presenting the error to the user in the nicely-formatted way introduced in this code, we seem to be implying that this is something the user needs to worry about. Yet, if this is a program error, we should say that this is a System Error and they need to contact Support for resolution. Note also that for unexpected program errors, there is a well-defined generic way we catch exceptions in the Fragment Executor. The Fragment Executor is the handler of last resort, will report an error, and will shut down the fragment. So, basic question: what is the source of the IOBE? If the IOBE is a program error, do we need the fancy handling if the Fragment Executor handles such errors? And, if the IOBE indicates a problem the user can resolve, should the error messages indicate a possible resolution? What might that problem be? Corrupted database? Incorrect SQL? Update done concurrently with a query? Something else?
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
