Yuriy Flyud created NIFI-6744:
---------------------------------
Summary: PutSolrContentStream Processor does no route to
connection_failure in Cloud mode if Zookeeper is not running
Key: NIFI-6744
URL: https://issues.apache.org/jira/browse/NIFI-6744
Project: Apache NiFi
Issue Type: Bug
Components: Core Framework
Affects Versions: 1.9.2
Reporter: Yuriy Flyud
Solr processors would not detect connection failure in case ZK is not running.
When this happens a SolrException is thrown. The following causes should route
a flow file to a connection_failure:
* java.util.concurren.TimeoutException - when ZK was down during Solr
processor start
* org.apache.zookeeper.KeeperException.ConnectionLossException - when ZK
stopped after Solr processor was started and connection established.
This could be fixed by implementing a SolrException handling in the following
way:
{code:java}
catch (SolrException e) {
//Additional handling of ZK connection exceptions.
if (causedByException(e, TimeoutException.class) || causedByException(e,
ConnectionLossException.class)) {
connectionError.set(e);
} else {
error.set(e);
}
...
private boolean causedByException(Exception e, Class<? extends Exception> ec) {
boolean foundException = false;
Throwable cause = e.getCause();
while (cause != null) {
if (ec.isInstance(cause)) {
foundException = true;
break;
}
cause = cause.getCause();
}
return foundException;
}{code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
