Prevent logging of stacktrace for bad Cassandra contact point

1
votes

When my Cassandra client program, which uses the Datastax Java driver, is given an invalid contact point (a hostname of a computer than is not actually running a Cassandra daemon) the driver itself logs a stacktrace. The stacktrace is worthless however, as there is a configuration error rather than a bug, and it is preceded by a much more informative warning message.

How can I configure the Cassandra driver not to trow an exception in this case, or configure logback not to log the stacktrace?

Here are the noisy log messages I get at present.:

2015-05-07 13:55:22,758 my-program: WARN You listed test-host-2.example.com/172.16.12.202:9042 in your contact points, but it could not be reached at startup
2015-05-07 13:55:22,919 my-program: WARN Some contact points don't match specified local data center. Local DC = DC1. Non-conforming contact points: /172.16.12.204:9042 (DC2)
2015-05-07 13:55:28,105 my-program: ERROR Error creating pool to test-host-2.example.com/172.16.12.202:9042
com.datastax.driver.core.TransportException: [test-host-2.example.com/172.16.12.202:9042] Cannot connect
    at com.datastax.driver.core.Connection.(Connection.java:106) ~[my-program-1.0.0.1.jar:1.0.0.1]
    at com.datastax.driver.core.PooledConnection.(PooledConnection.java:32) ~[my-program-1.0.0.1.jar:1.0.0.1]
    at com.datastax.driver.core.Connection$Factory.open(Connection.java:521) ~[my-program-1.0.0.1.jar:1.0.0.1]
    at com.datastax.driver.core.SingleConnectionPool.(SingleConnectionPool.java:76) ~[my-program-1.0.0.1.jar:1.0.0.1]
    at com.datastax.driver.core.HostConnectionPool.newInstance(HostConnectionPool.java:35) ~[my-program-1.0.0.1.jar:1.0.0.1]
    at com.datastax.driver.core.SessionManager.replacePool(SessionManager.java:239) ~[my-program-1.0.0.1.jar:1.0.0.1]
    at com.datastax.driver.core.SessionManager.access$400(SessionManager.java:39) ~[my-program-1.0.0.1.jar:1.0.0.1]
    at com.datastax.driver.core.SessionManager$3.call(SessionManager.java:272) [my-program-1.0.0.1.jar:1.0.0.1]
    at com.datastax.driver.core.SessionManager$3.call(SessionManager.java:264) [my-program-1.0.0.1.jar:1.0.0.1]
    at java.util.concurrent.FutureTask.run(FutureTask.java:262) [na:1.7.0_75]
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145) [na:1.7.0_75]
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615) [na:1.7.0_75]
    at java.lang.Thread.run(Thread.java:745) [na:1.7.0_75]
Caused by: org.jboss.netty.channel.ConnectTimeoutException: connection timed out: test-host-2.example.com/172.16.12.202:9042
    at org.jboss.netty.channel.socket.nio.NioClientBoss.processConnectTimeout(NioClientBoss.java:137) ~[my-program-1.0.0.1.jar:1.0.0.1]
    at org.jboss.netty.channel.socket.nio.NioClientBoss.process(NioClientBoss.java:83) ~[my-program-1.0.0.1.jar:1.0.0.1]
    at org.jboss.netty.channel.socket.nio.AbstractNioSelector.run(AbstractNioSelector.java:318) ~[my-program-1.0.0.1.jar:1.0.0.1]
    at org.jboss.netty.channel.socket.nio.NioClientBoss.run(NioClientBoss.java:42) ~[my-program-1.0.0.1.jar:1.0.0.1]
    ... 3 common frames omitted
2
cassandralogbackdatastax-java-driver

2 Answers

0
votes

This sounds like a feature request. Feel free to create a jira - https://datastax-oss.atlassian.net/secure/Dashboard.jspa

You could turn down logging but I don't think you want to exclude ERRORs or connection time outs.

Are you just bothered by the ERRORs in your logs? It may be useful to know when you have downed nodes that are contact points...

0
votes

In the general case it makes sense to show the stack trace, it could be a different error (e.g. the server does have Cassandra running but authentication is enabled and you're not providing the right credentials).

If you really want to suppress stack traces in Logback, this is apparently possible with a custom layout.