[jira] [Commented] (THRIFT-1323) TFramedTransport should throw an exception that distinguishes an oversized frame from a dead client, servers should log that exception

["Bryan Duxbury (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/THRIFT-1323?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13098422#comment-13098422 ] 

Bryan Duxbury commented on THRIFT-1323:
---------------------------------------

I think it would be hard to make this any other exception than a TTransportException, though we could make a subclass if that would be more helpful. As far as just swallowing the exception, I guess that's really up to the implementing server. You could make the argument that you should never swallow an exception. I would commit a patch that unconditionally logged all exceptions, especially if it was at a log level that was easier to silence if the messages were too superfluous.

> TFramedTransport should throw an exception that distinguishes an oversized frame from a dead client, servers should log that exception
> --------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: THRIFT-1323
>                 URL: https://issues.apache.org/jira/browse/THRIFT-1323
>             Project: Thrift
>          Issue Type: Bug
>          Components: Java - Library
>            Reporter: Jim Ancona
>
> When TFramedTransport receives a frame that is larger than the maximum frame size, it throws a TTransportException. This makes that error indistinguishable from a failed client. Because such errors are common, many servers (TSimpleServer, TThreadPoolServer, and Cassandra's CustomTThreadPoolServer) swallow TTransportException without a log message, making it difficult to diagnose this problem. I'm not familiar enough with the code base to suggest which exception to throw in its place, although a subclass of TTransportException, or a new TTransportException.type_ value might work. In any case, the corresponding server implementations should log that condition when it occurs.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira