[
https://issues.apache.org/jira/browse/BEAM-6886?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16820820#comment-16820820
]
Egbert commented on BEAM-6886:
------------------------------
[~echauchot] I'm not sure if I can find the time to contribute a patch, I'll
keep it in mind.
For BigQueryIO I actually meant there is a GroupByKey operation in the write. I
think it's intended for deduplication of records, but as a side-effect it seems
to batch together windows in multiple bundles. It's not something you have to
do explicitly anyway. The same holds for PubsubIO#writeMessages - that also
collects multiple messages in a short time period to be published in one batch.
I didn't dive into the code to see how it is done there, though.
> Change batch handling in ElasticsearchIO to avoid necessity for
> GroupIntoBatches
> --------------------------------------------------------------------------------
>
> Key: BEAM-6886
> URL: https://issues.apache.org/jira/browse/BEAM-6886
> Project: Beam
> Issue Type: Improvement
> Components: io-java-elasticsearch
> Affects Versions: 2.11.0
> Reporter: Egbert
> Priority: Major
>
> I have a streaming job inserting records into an Elasticsearch cluster. I set
> the batch size appropriately big, but I found out this is not causing any
> effect at all: I found that all elements are inserted in batches of 1 or 2
> elements.
> The reason seems to be that this is a streaming pipeline, which may result in
> tiny bundles. Since ElasticsearchIO uses `@FinishBundle` to flush a batch,
> this will result in equally small batches.
> This results in a huge amount of bulk requests with just one element,
> grinding the Elasticsearch cluster to a halt.
> I have now been able to work around this by using a `GroupIntoBatches`
> operation before the insert, but this results in 3 steps (mapping to a key,
> applying GroupIntoBatches, stripping key and outputting all collected
> elements), making the process quite awkward.
> A much better approach would be to internalize this into the ElasticsearchIO
> write transform.. Use a timer that flushes the batch at batch size or end of
> window, not at the end of a bundle.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
