[jira] [Updated] (KAFKA-14274) Introduce FetchRequestManager

["Kirk True (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/KAFKA-14274?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Kirk True updated KAFKA-14274:
------------------------------
    Summary: Introduce FetchRequestManager  (was: Split Fetcher into Fetcher and MetadataFetcher)

> Introduce FetchRequestManager
> -----------------------------
>
>                 Key: KAFKA-14274
>                 URL: https://issues.apache.org/jira/browse/KAFKA-14274
>             Project: Kafka
>          Issue Type: Sub-task
>          Components: clients, consumer
>            Reporter: Philip Nee
>            Assignee: Kirk True
>            Priority: Major
>
> The {{Fetcher}} class is used internally by the {{KafkaConsumer}} to fetch records from the brokers. There is ongoing work to create a new consumer implementation with a significantly refactored threading model. The threading refactor work requires a similarly refactored {{{}Fetcher{}}}.
> The fetch request and fetch processing should happen asynchronously. More specifically, we have the background thread to send fetch requests autonomously and relay the response back to the foreground thread. The foreground thread collects these fetch responses in the form of {{{}CompletedFetch{}}}, decompresses the data, deserializes the data, and converts each {{Record}} into a {{ConsumerRecord}} for returning to the user.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)