[
https://issues.apache.org/jira/browse/FLINK-34554?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17824757#comment-17824757
]
Hilmi Al Fatih edited comment on FLINK-34554 at 3/8/24 11:49 AM:
-----------------------------------------------------------------
{quote}You're not using the latest version of the Flink Kafka connector. The
one that's bundled with Flink 1.17 is outdated. Since Flink connectors have
been externalized, you'll need to make sure that you're updating to the latest
version yourself. For the Flink Kafka connector, that's currently 3.1.0-1.17
{quote}
I see. I thought I can just use the connector bundled with flink 1.17.
Btw do you think it will be stable to keep using FlinkKafkaProducer (bundled
with connector 3.1.0) with flink 1.17? If so I would like to proceed this way
since I believe if I use KafkaSink (even with connector 3.1.0), the same issue
will happen again since its underlying implementation is generally the same
(recreate transactionId per checkpoint).
{quote}IIRC it was a deliberate choice to not reusing the pooling, but I can't
quickly find anymore why that was the case. That the implementation differs
isn't too surprising: FlinkKafkaProducer used the SinkFunction, while KafkaSink
uses Flink's unified Sink API.
{quote}
If you could somehow find reference on why pooling was not chosen for the new
KafkaSink, I would love to know!
was (Author: JIRAUSER283529):
{quote}You're not using the latest version of the Flink Kafka connector. The
one that's bundled with Flink 1.17 is outdated. Since Flink connectors have
been externalized, you'll need to make sure that you're updating to the latest
version yourself. For the Flink Kafka connector, that's currently 3.1.0-1.17
{quote}
I see. I thought I can just use the connector bundled with flink 1.17.
Btw do you think it will be stable to keep using FlinkKafkaProducer (bundled
with connector 3.1.0) with flink 1.17? If so I would like to proceed this way
since I believe if I use KafkaSink (even with connector 3.1.0), the same issue
will happen again since its underlying implementation is generally the same.
{quote}IIRC it was a deliberate choice to not reusing the pooling, but I can't
quickly find anymore why that was the case. That the implementation differs
isn't too surprising: FlinkKafkaProducer used the SinkFunction, while KafkaSink
uses Flink's unified Sink API.
{quote}
If you could somehow find reference on why pooling was not chosen for the new
KafkaSink, I would love to know!
> Using EXACTLY_ONCE with KafkaSink cause broker's OOM due to newly created
> transactionalId per checkpoint
> --------------------------------------------------------------------------------------------------------
>
> Key: FLINK-34554
> URL: https://issues.apache.org/jira/browse/FLINK-34554
> Project: Flink
> Issue Type: Improvement
> Components: Connectors / Kafka
> Affects Versions: 1.16.3, 1.17.2, 1.18.1
> Reporter: Hilmi Al Fatih
> Priority: Major
> Attachments: image (4).png, image (5).png
>
>
> Flink version: 1.17.1
> Kafka Broker version: 2.7.1 * 4 GB heap memory for each
> Hi, We recently had an outage in our production system after we perform a
> Flink kafka-connector API upgrade. To give a brief context, our application
> is a simple kafka-to-kafka pipeline with minimal processing. We run in
> EXACTLY_ONCE mode, thus kafka transaction is involved.
> Our application runs with total around 350 sink subtask. Checkpoint period
> was set to 5 seconds to avoid blocking {{read_committed}} consumers too long.
> We recently performed an upgrade with the following details:
> Previous state:
> * Flink version: 1.14.4
> * Broker version: 2.7.1
> * kafka connector API: FlinkKafkaProducer
> Update to:
> * Flink version: 1.17.1
> * Broker version: 2.7.1
> * kafka connector API: KafkaSink
> Around 10 hours after the deployment, our kafka broker started to failing
> with OOM error. Heap dump entries are dominated by the ProducerStateEntry
> records.
> Our investigation leads to finding the total implementation change between
> FlinkKafkaProducer and KafkaSink.
> * KafkaSink generate different transactionalId for each checkpoint,
> * FlinkKafkaProducer uses constant set of transactionalId pool.
> With this behavior, KafkaSink seemed to exhaust our broker heap very fast and
> the ProducerStateEntry will only expire after
> [transactional.id.expiration.ms|http://transactional.id.expiration.ms/] ,
> which by default is set to 7 days.
> ([ref1|https://github.com/apache/kafka/blob/61dbce85d0d41457d81a4096ecaea049f3a4b3ae/core/src/main/scala/kafka/log/Log.scala#L677],
>
> [ref2|https://github.com/apache/kafka/blob/61dbce85d0d41457d81a4096ecaea049f3a4b3ae/core/src/main/scala/kafka/log/LogManager.scala#L268],
>
> [ref3|https://github.com/apache/kafka/blob/61dbce85d0d41457d81a4096ecaea049f3a4b3ae/core/src/main/scala/kafka/log/LogManager.scala#L1207])
> For our job, it means it creates roughly:
> * 10 hour running) 350 ids * 12 times/minute * 60 min/hour * 10 hour ~
> 2,520,000
> * 7 days) ~ 42mil entries.
> Attached below is the number of ProducerStateEntry entries of heap dump when
> it is OOM:
> * 505,000 (6.5%), in total it would be roughly ~ 7,000,000 entries.
> There are several things that come up in our mind to mitigate the drawbacks
> such as:
> * reduce the number of subtasks, so it reduces the number of transactionalId
> * Enlarge the checkpoint period to reduce the newly generated
> transactionalId rate.
> * Shorten
> [transactional.id.expiration.ms|http://transactional.id.expiration.ms/] to
> expire the unused transactionalId soon.
> * Increase the broker heap
> However, above mitigation might be too cumbersome and need careful tuning
> which harm our flexibility.In addition, due to the lack of maintaining
> lingering transaction state, TransactionAborter seems to abort old
> transaction naively. We might be accidentally (or purposefully) reuse the
> same transactionalIdPrefix and start the counter from 0. In that case, if the
> old transactionalId happens to have epoch >0, it will keep looping aborting
> the nonexistent transactions up to the latest checkpoint counter (which may
> be too big) and make the job stuck.
> Btw, I am aware that in Flink 2.0, you guys are putting a lot of effort on
> creating better integration with Kafka transaction
> ([FLIP-319|https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=255071710]).
> In FLIP-319, it mentions something about TID pooling. However, it is seem
> that there is no relevant page yet for it, so I wonder whether there are any
> concrete plan already that I can follow, or if there is something I can
> contribute to, I will be really happy to help.
>
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
