[
https://issues.apache.org/jira/browse/HADOOP-19609?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18008596#comment-18008596
]
ASF GitHub Bot commented on HADOOP-19609:
-----------------------------------------
bhattmanish98 opened a new pull request, #7817:
URL: https://github.com/apache/hadoop/pull/7817
JIRA: https://issues.apache.org/jira/browse/HADOOP-19609
**Description**
The ABFS (Azure Blob File System) driver currently utilizes the default JDK
HTTP client to communicate with the Azure Storage backend service, known as
XFE. While the JDK client is lightweight, simple, and built into the JDK, it
falls short when it comes to providing advanced connection management
capabilities.
By switching to Apache HTTP Client, the ABFS driver aims to gain better
control over HTTP connections while enabling enterprise-level features such as
configurable keep-alive strategies, idle connection eviction, retries on
failure, and better resource usage tracking.
This is the initial PR where we have optimized the structure of keep alive
cache for Apache client. Migration from JDK and Apache client will be done in
subsequent PR.
> ABFS: Apache Client Connection Pool Relook
> ------------------------------------------
>
> Key: HADOOP-19609
> URL: https://issues.apache.org/jira/browse/HADOOP-19609
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: fs/azure
> Affects Versions: 3.4.1
> Reporter: Manish Bhatt
> Assignee: Manish Bhatt
> Priority: Major
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
