[jira] [Commented] (FLINK-16057) Performance regression in ContinuousFileReaderOperator

["Aljoscha Krettek (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/FLINK-16057?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17084694#comment-17084694 ] 

Aljoscha Krettek commented on FLINK-16057:
------------------------------------------

[~pnowojski] [~roman_khachatryan] Is there any more work happening on this one?

> Performance regression in ContinuousFileReaderOperator
> ------------------------------------------------------
>
>                 Key: FLINK-16057
>                 URL: https://issues.apache.org/jira/browse/FLINK-16057
>             Project: Flink
>          Issue Type: Bug
>          Components: API / DataStream
>    Affects Versions: 1.11.0
>            Reporter: Roman Khachatryan
>            Assignee: Roman Khachatryan
>            Priority: Blocker
>              Labels: pull-request-available
>             Fix For: 1.11.0
>
>          Time Spent: 20m
>  Remaining Estimate: 0h
>
> After switching CFRO to a single-threaded execution model performance regression was expected to be about 15-20% (benchmarked in November).
> But after merging to master it turned out to be about 50%.
>   
> One reason is that the chaining strategy isn't set by default in CFRO factory.
> Without that even reading and outputting all records of a split in a single mail action doesn't reverse the regression (only about half).
> However,  with strategy set AND batching enabled fixes the regression (starting from batch size 6).
> Though batching can't be used in practice because it can significantly delay checkpointing.
>  
> Another approach would be to process one record and the repeat until defaultMailboxActionAvailable OR haveNewMail.
> This reverses regression and even improves the performance by about 50% compared to the old version.
>  
> The final solution could also be FLIP-27.
>  
> Other things tried (didn't help):
>  * CFRO rework without subsequent commits (removing checkpoint lock)
>  * different batch sizes, including the whole split, without chaining strategy fixed - partial improvement only
>  * disabling close
>  * disabling checkpointing
>  * disabling output (serialization)
>  * using LinkedList instead of PriorityQueue
>  



--
This message was sent by Atlassian Jira
(v8.3.4#803005)