> 发送时间: 2021年5月24日 23:17
> 收件人: Yangze Guo
> 抄送: ro...@apache.org ; user
> 主题: Re: ES sink never receive error code
>
> Got it! thanks for helping.
>
> On Thu, May 20, 2021 at 7:15 PM Yangze Guo wrote:
>
> > So, ES BulkProcessor retried after bulk request was pa
he thread. Therefore
please upgrade to connector 7.*.
发件人: Qihua Yang
发送时间: 2021年5月24日 23:17
收件人: Yangze Guo
抄送: ro...@apache.org ; user
主题: Re: ES sink never receive error code
Got it! thanks for helping.
On Thu, May 20, 2021 at 7:15 PM Yangze Guo
mailt
Got it! thanks for helping.
On Thu, May 20, 2021 at 7:15 PM Yangze Guo wrote:
> > So, ES BulkProcessor retried after bulk request was partially rejected.
> And eventually that request was sent successfully? That is why failure
> handler was not called?
>
> If the bulk request fails after the max
> So, ES BulkProcessor retried after bulk request was partially rejected. And
> eventually that request was sent successfully? That is why failure handler
> was not called?
If the bulk request fails after the max number of retries
(bulk.flush.backoff.retries), the failure handler will still be
c
Thank you for the reply!
Yes, we did config bulk.flush.backoff.enable.
So, ES BulkProcessor retried after bulk request was partially rejected. And
eventually that request was sent successfully? That is why failure handler
was not called?
Thanks,
Qihua
On Thu, May 20, 2021 at 2:22 PM Roman Khachat
Hi,
Have you tried to change bulk.flush.backoff.enable?
According to the docs [1], the underlying ES BulkProcessor will retry
(by default), so the provided failure handler might not be called.
[1]
https://ci.apache.org/projects/flink/flink-docs-stable/docs/connectors/datastream/elasticsearch/#con
Hello,
We are using flink-connector-elasticsearch6_2.11 to ingest stream data to
ES by using bulk requests. From ES metrics, we observed some bulk thread
pool rejections. Contacted AWS team, their explanation is part of bulk
request was rejected. Response body should include status for each item.
F
