dlmarion commented on issue #2700: URL: https://github.com/apache/accumulo/issues/2700#issuecomment-1136438504
> The default is 5000, not 1000. You're right, I conflated that with ROUNDS_MIN. > But the question remains: do we want fast and efficient checks, or do we want slow checks/large iterations? I get the rationale behind the slow checks / large iterations. I don't know that the lock contention is intentional as a throttle mechanism. I have no issue with that happening the first time to make sure that the password passed in is in fact the correct one. I think we need to do something here based on @keith-turner 's comment in #2707 (regarding the difference between caching and not caching successful password checks under load): ``` With these changes and 500 concurrent queries the average time is 400ms and the max times are around 1000ms (ignoring the first few lines because JIT probably has not kicked in scan servers). Without these changes the average times are around 1600ms to 1700ms and the max times 15,000ms to 20,000ms ``` I can't imagine that this is the desired behavior. Was it this bad before the change to commons-codec? -- 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]
