andrewglowacki commented on issue #2331: URL: https://github.com/apache/accumulo/issues/2331#issuecomment-953331475
Sorry, I must be missing something, or I didn't explain the situation well enough. The tablet server itself is throwing an OOM error which is sent back to the client over the thrift RPC. The client itself is not OOM. The existing behavior causes the batch writer to get stuck indefinitely. The binning thread should alert the user code that an error occurred so that it can retry with a new batch writer. A single bad tablet server should not bring a client writer to halt. Similarly, when any kind of Throwable occurs in a batch writer SendTask, the task properly catches the Throwable and alerts the user code. The same should be done during mutation binning. -- 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]
