But the buffering can sometimes surprise you. Write combining buffers (on atom, core, core duo, core solo, P4) hold modified cache lines before the stores are complete. flushing the buffer cache would cause you to wait for every block with a modification to be written out to disk. 3. buffer flush took longer time than slow_flush_log_threshold: plugin_id=“object:1d8281c” elapsed_time=227.446416143 slow_flush_log_threshold=20.0 刷新日志的时间太长了，解决办法，就是减小 flush_interval的时间。 This is called flushing the buffer. Red Hat OpenShift Container Platform 3.x Here the performance gain is even bigger than for reading, about 60%. Suppose you have a program like this: 2017-09-25 16:07:41 +0200 [warn]: buffer flush took longer time than slow_flush_log_threshold: plugin_id="object:13c4370" elapsed_time=36.999487413 slow_flush_log_threshold=20.0 2017-09-25 16:07:38 +0200 [warn]: Could not push logs to Elasticsearch, resetting … When this happens, the buffer was empty and the returned frame is up to date. In this case, LRU flushing is adjusted to be one (1) percent more aggressive for the specific bufferpool where the page replacement using high priority buffer took place. It takes a long time to flush buffer. 2019-01-09 16: 37: 59 + 0800 [warn]: # 16 buffer flush took longer time than slow_flush_log_threshold: elapsed_time= 21.684925531037152 slow_flush_log_threshold= 20.0 plugin_id= " object:3fb71aba75a0 " For instance, stores happen after the instruction retires and the store buffer keeps hold the address and data until the store retires. These can take a relatively long time to complete. Surprise! 2019-04-27 03:24:01 +0900 [info]: #0 flushing all buffer forcedly 2019-04-27 23:00:53 +0900 [warn]: #0 buffer flush took longer time than slow_flush_log_threshold: elapsed_time=26.38604311645031 slow_flush_log_threshold=20.0 plugin_id="object:3fdfe37a9ed4" 2019-04-28 03:33:01 +0900 [info]: #0 force flushing buffered events Continue querying frames until the time measured exceeds a certain limit. When the buffer is full, Perl writes all the data in the buffer at once. Instead, it goes into a buffer. why do you say that? In case you have (already) huge log files, this will result in slower booting. (The answer on the linked post shows: returning a frame from the buffer takes about 1/8th the time of returning an … The systemd-journal-flush.service asks the journal daemon to flush any log data stored in /run/log/journal into /var/log/journal, if persistent storage is enabled. Many warning raise in fluentd log 2019-04-29 00:23:33 +0000 [warn]: buffer flush took longer time than slow_flush_log_threshold: elapsed_time=31.253766102716327 slow_flush_log_threshold=20.0 plugin_id="object:3fb36e8d4058" 2019-04-29 00:23:33 +0000 [warn]: buffer flush took longer time than slow_flush_log_threshold: elapsed_time=31.261768959928304 slow_flush_log_threshold=20.0 … It should take <5ms or do you mean "flush the ENTIRE buffer cache", in which case - it could take 5ms * number of blocks needed to be written to disk. Buffer flush took longer time messages; 2019-09-25 01:45:26 -0700 [warn]: buffer flush took longer time than slow_flush_log_threshold: plugin_id="object:3ea119ad7d58" elapsed_time=22.682930512 slow_flush_log_threshold=20.0 Environment.