[
https://issues.apache.org/jira/browse/DRILL-4523?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15220696#comment-15220696
]
ASF GitHub Bot commented on DRILL-4523:
---------------------------------------
Github user parthchandra commented on a diff in the pull request:
https://github.com/apache/drill/pull/445#discussion_r58130327
--- Diff:
exec/java-exec/src/main/java/org/apache/drill/exec/service/ServiceEngine.java
---
@@ -177,6 +180,16 @@ public void run() {
});
}
+ /** If we are in distributed mode, disallow using loopback address to
register new drillbit
+ * unless zookeeper is on the same node with drillbit. */
+ private void checkLoopbackAddress(String address) throws
DrillbitStartupException, UnknownHostException {
+ String coordAddress =
StringUtils.substringBefore(config.getString(ExecConstants.ZK_CONNECTION), ":");
--- End diff --
At this point, how do you know you're in distributed and not embedded mode?
> Disallow using loopback address in distributed mode
> ---------------------------------------------------
>
> Key: DRILL-4523
> URL: https://issues.apache.org/jira/browse/DRILL-4523
> Project: Apache Drill
> Issue Type: Improvement
> Components: Server
> Affects Versions: 1.6.0
> Reporter: Arina Ielchiieva
> Assignee: Arina Ielchiieva
> Fix For: 1.7.0
>
>
> If we enable debug for org.apache.drill.exec.coord.zk in logback.xml, we only
> get the hostname and ports information. For example:
> {code}
> 2015-11-04 19:47:02,927 [ServiceCache-0] DEBUG
> o.a.d.e.c.zk.ZKClusterCoordinator - Cache changed, updating.
> 2015-11-04 19:47:02,932 [ServiceCache-0] DEBUG
> o.a.d.e.c.zk.ZKClusterCoordinator - Active drillbit set changed. Now
> includes 2 total bits. New active drillbits:
> h3.poc.com:31010:31011:31012
> h2.poc.com:31010:31011:31012
> {code}
> We need to know the IP address of each hostname to do further troubleshooting.
> Imagine if any drillbit registers itself as "localhost.localdomain" in
> zookeeper, we will never know where it comes from. Enabling IP address
> tracking can help this case.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)