[elasticsearch] buffer flush took longer time than slow_flush_log_threshold

[wael ghaith]

fluentd logs :

2020-02-19 08:18:51 +0000 [warn]: [elasticsearch] buffer flush took longer time than slow_flush_log_threshold: elapsed_time=38.35212400369346 slow_flush_log_threshold=35.0 plugin_id="elasticsearch"
2020-02-19 08:18:56 +0000 [warn]: [elasticsearch] buffer flush took longer time than slow_flush_log_threshold: elapsed_time=48.19527011923492 slow_flush_log_threshold=35.0 plugin_id="elasticsearch"
2020-02-19 08:19:00 +0000 [warn]: [elasticsearch] buffer flush took longer time than slow_flush_log_threshold: elapsed_time=81.41292346082628 slow_flush_log_threshold=35.0 plugin_id="elasticsearch"
2020-02-19 08:19:11 +0000 [warn]: [elasticsearch] buffer flush took longer time than slow_flush_log_threshold: elapsed_time=61.737419448792934 slow_flush_log_threshold=35.0 plugin_id="elasticsearch"

What is the cause of this problem?
What are the pods the Fluentd connects with it? (client or master or data)

[Akhisar]

In my experience, I would ask you to check on the resource utilization(mainly CPU) on the elasticsearch cluster. Can you also please check on the elastic logs?