[
https://issues.apache.org/jira/browse/BLUR-54?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13592969#comment-13592969
]
Aaron McCurry commented on BLUR-54:
-----------------------------------
I believe the problem may be an uncaught non BlurException, because if a
BlurException is thrown the client will not retry the call. If the NPE
exception is thrown back through thrift on the server-side then a TException is
thrown to the client and that will cause bad connection / retries to occur. So
I think that this is actually 2 problems, one we should provide a better
exception and two we should not allow non-BlurExceptions to bubble up passed
the BlurServer api, basically catch everything and wrap in a BlurException (See
BException as a helper class).
Aaron
> Improve the error when a bad shard index is provided.
> -----------------------------------------------------
>
> Key: BLUR-54
> URL: https://issues.apache.org/jira/browse/BLUR-54
> Project: Apache Blur
> Issue Type: Bug
> Affects Versions: 0.2.0
> Reporter: Tim Williams
> Priority: Trivial
>
> Currently, if a client passes an unknown shard index, they get a message
> about a bad connection string when in fact that's just a side effect. I'm
> thinking that the TableLayout implemented in ThriftBlurServer should throw a
> RuntimeException if it isn't found - because that's where we have some useful
> context. Sound reasonable? Another option is to throw it in BlurServer's
> getConnection.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
