[
https://issues.apache.org/jira/browse/FLINK-14845?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16979249#comment-16979249
]
Piotr Nowojski commented on FLINK-14845:
----------------------------------------
[~kevin.cyj], I think I would prefer the 3rd option, but I think you are right,
potential implementation cost of this option might not be worth it for now.
I guess the difference between 2. and 4. in terms of an implementation is very
small? We could start with 4. and re-evaluate 3., if we ever get such request
from a user.
Side note, [~AHeise] made me aware that we might find this compression useful
also for spilling the buffers for unaligned checkpoints.
Also just for the record. I was thinking about another solution, to
compress/decompress the data in Netty handlers. For streaming that would solve
all of the problems with network flushes and small records/buffers (Netty would
be compressing multiplexed channels), but again that wouldn't be very helpful
for the {{BoundedBlockingSubpartition}}. Again, for now we can stick with the
common compression code path shared between streaming/batch and if this ever
becomes an issue, we can re-visit.
> Introduce data compression to blocking shuffle.
> -----------------------------------------------
>
> Key: FLINK-14845
> URL: https://issues.apache.org/jira/browse/FLINK-14845
> Project: Flink
> Issue Type: Sub-task
> Components: Runtime / Network
> Reporter: Yingjie Cao
> Assignee: Yingjie Cao
> Priority: Major
>
> Currently, blocking shuffle writer writes raw output data to disk without
> compression. For IO bounded scenario, this can be optimized by compressing
> the output data. It is better to introduce a compression mechanism and offer
> users a config option to let the user decide whether to compress the shuffle
> data. Actually, we hava implemented compression in our inner Flink version
> and here are some key points:
> 1. Where to compress/decompress?
> Compressing at upstream and decompressing at downstream.
> 2. Which thread do compress/decompress?
> Task threads do compress/decompress.
> 3. Data compression granularity.
> Per buffer.
> 4. How to handle that when data size become even bigger after compression?
> Give up compression in this case and introduce an extra flag to identify if
> the data was compressed, that is, the output may be a mixture of compressed
> and uncompressed data.
>
> We'd like to introduce blocking shuffle data compression to Flink if there
> are interests.
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
