[jira] [Comment Edited] (KAFKA-3927) kafka broker config docs issue

["Shawn Guo (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/KAFKA-3927?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15363608#comment-15363608 ] 

Shawn Guo edited comment on KAFKA-3927 at 7/6/16 2:25 AM:
----------------------------------------------------------

these parameters have default value in 0.8 version in config docs(as below).
https://kafka.apache.org/08/configuration.html
log.flush.interval.ms	3000
log.flush.scheduler.interval.ms	3000
log.flush.interval.messages	10000

For latest version, both log.flush.interval.ms and log.flush.interval.messages become passive and no explicit default value.  Does it mean that official recommends not to set these and use replication for durability and allow the operating system's background flush capabilities as it is more efficient?

If it is true, I will close this issue. maybe I have a wrong understanding of recent durability strategy changes.


was (Author: guoxu1231):
these parameters have default value in 0.8 version in config docs, " Logs are flushed when they get rotated based on other configs such as log.segment.size etc.. ". this sounds crazy idea.. 
https://kafka.apache.org/08/configuration.html
log.flush.interval.ms	3000
log.flush.scheduler.interval.ms	3000
log.flush.interval.messages	10000

> kafka broker config docs issue
> ------------------------------
>
>                 Key: KAFKA-3927
>                 URL: https://issues.apache.org/jira/browse/KAFKA-3927
>             Project: Kafka
>          Issue Type: Bug
>          Components: website
>    Affects Versions: 0.10.0.0
>            Reporter: Shawn Guo
>            Priority: Minor
>
> https://kafka.apache.org/documentation.html#brokerconfigs
> log.flush.interval.messages 
> default value is "9223372036854775807"
> log.flush.interval.ms 
> default value is null
> log.flush.scheduler.interval.ms 
> default value is "9223372036854775807"
> etc. obviously these default values are incorrect. how these doc get generated ? it looks confusing. 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)