In terms of avoiding collapse Logger waiting for the sent of messages to Graylog, I've developed another implementation based on the sending on the messages through workers in an async way. This is completely backwards compatible and even fixes some issues regarding tests.
You can see the changes in README to see what's changed. Mainly it's only based on the addition of the async backend Logger.Backends.GelfAsync and the pool_size configuration parameter.
In terms of avoiding collapse Logger waiting for the sent of messages to Graylog, I've developed another implementation based on the sending on the messages through workers in an async way. This is completely backwards compatible and even fixes some issues regarding tests.
You can see the changes in README to see what's changed. Mainly it's only based on the addition of the async backend
Logger.Backends.GelfAsyncand thepool_sizeconfiguration parameter.