[
https://issues.apache.org/jira/browse/HIVE-18389?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16353696#comment-16353696
]
Peter Vary commented on HIVE-18389:
-----------------------------------
[~stakiar]: I have questions about this part of the code:
{code:java}
private void handle(ChannelHandlerContext ctx, SparkUIWebURL msg) {
String printMsg = "Hive on Spark Session Web UI URL: " +
msg.UIWebURL;
consoleStream.println(printMsg);
LOG.info(printMsg);
}
{code}
The goal of this change to print the Web UI URL to the logs, and to the console
as well? If I remember correctly, if we use the {{LogHelper.logInfo}} we can
archive the same in one line, with the additional benefit of honoring if the
client set the output to {{silent}}. So if we use {{LogHelper console}} instead
of {{PrintStream consoleStream}} we might better of in the end.
Is there any drawback with this approach?
Thanks,
Peter
> Print out Spark Web UI URL to the console log
> ---------------------------------------------
>
> Key: HIVE-18389
> URL: https://issues.apache.org/jira/browse/HIVE-18389
> Project: Hive
> Issue Type: Sub-task
> Components: Spark
> Reporter: Sahil Takiar
> Assignee: Sahil Takiar
> Priority: Major
> Attachments: HIVE-18389.1.patch
>
>
> Should be accessible via {{SparkContext#uiWebUrl}}. It just needs to be sent
> from the {{RemoteDriver}} to HS2.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
