[
https://issues.apache.org/jira/browse/THRIFT-3769?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16437733#comment-16437733
]
Vihang Karajgaonkar commented on THRIFT-3769:
---------------------------------------------
Moving this forward. When load balancers like HAProxy are used for HiveServer2
these exceptions practically make the HS2 logs unusable since it logs
connection resets every 2 sec. I used patch provided by [~qwertymaniac] to
create a PR. [~sershe] or [~jking3] Can you please review? Thanks!
> Fix logic of THRIFT-2268
> ------------------------
>
> Key: THRIFT-3769
> URL: https://issues.apache.org/jira/browse/THRIFT-3769
> Project: Thrift
> Issue Type: Bug
> Components: Java - Library
> Affects Versions: 0.9.2
> Reporter: Harsh J
> Priority: Minor
> Attachments: THRIFT-3769.patch
>
>
> THRIFT-2268 intended to reduce the logging noise of TSaslTransport enabled
> servers, but the commit doesn't help in doing so (its just adding more
> specific noise today).
> This is because the transport factory overrides disallow throwing specific
> execution types (no TTransportException in signature), and thereby all
> implementations will throw a RuntimeException-wrapped exception, which the
> added catch clauses would never encounter.
> https://github.com/apache/thrift/blob/master/lib/java/src/org/apache/thrift/transport/TSaslServerTransport.java#L217-L219
> https://github.com/apache/thrift/blob/master/lib/java/src/org/apache/thrift/server/TThreadPoolServer.java#L290-L295
> We'll need to unwrap the RuntimeException to perform an actual valid cause
> check.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
