ramanathan1504 opened a new issue, #4369: URL: https://github.com/apache/logging-log4j2/issues/4369
`AbstractDatabaseManager.flush()` discards the buffered events when connecting, writing or committing fails (#4318, #4319). That bounds memory, but events the database never saw are lost: with `autoCommit=false` a failed commit rolls back, and a failed connect never sends anything. An alternative is to keep the batch and cap the buffer at `bufferSize`, so a short outage ends with the events written once the connection returns. That needs a policy decision: which event is dropped when the buffer is full, the oldest or the newest, and whether that is configurable. Raised on #4319, where the contributor proposed capping. Kept separate so the leak fix stays backportable. -- 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]
