Masahiro Nakagawa

@qingling128 I see. Does your record have specific field for your metadata? In v1, fluentd decide target chunk by chunk keys, ``. So if one, or more, field refers the...

> When I checked the logs of fluentd I saw that it is full of backslashes: Is this log single line, right? If so, it seems several logs are merged...

Just added new envvar to fluentd-kubernetes-daemonset for this case: https://github.com/fluent/fluentd-kubernetes-daemonset#use-fluent_container_tail_exclude_path-to-exclude-specific-container-logs

For lastest images, use cri parser is better than regexp: https://github.com/fluent/fluentd-kubernetes-daemonset#use-cri-parser-for-containerdcri-o-logs

Hmm.... looks like in_forward stops event loop to shutdown and it causes pending requests are ignored when at-most-once. Need to check event loop handling.

Very weird. https://github.com/fluent/fluentd/blob/d6b2c505f8c59661bbeef30fa590530f2dda1fe9/lib/fluent/plugin/in_forward.rb#L129 Here is the logging code. fluent-logger should send events as an array object like `[tag, time, record]`. But from your log, your application send integer object to...

> , the log messages have disappeared and td-agent is stable again just forwarding the log-tailing that remains, It means you use two input plugin, tail and forward, right? >...

Thanks! I will try to reproduce this problem with similar log volume.

I am now testing fluent-logger-ruby and in_forward combination with 300 logs/sec traffic since tomorrow but there is no error. Does your app send logs to in_forward concurrently? N loggers ->...

> Multiple threads on the same machine Sharing one fluent-logger between threads or each thread has each fluent-logger?