> On April 21, 2014, 5:08 p.m., Josh Elser wrote: > > test/src/test/java/org/apache/accumulo/test/functional/MonitorLoggingIT.java, > > line 48 > > <https://reviews.apache.org/r/20465/diff/2/?file=562356#file562356line48> > > > > Starting the Monitor before the cluster starts mind eliminate some > > churn. I forget if the tserver just has a watcher set to see the update or > > if it polls periodically. > > Bill Havanki wrote: > I'll give that a go.
The test wasn't happy with trying to start the monitor before the cluster. I would get either a NoNodeException or the monitor would apparently not remain up. > On April 21, 2014, 5:08 p.m., Josh Elser wrote: > > test/src/test/java/org/apache/accumulo/test/functional/MonitorLoggingIT.java, > > line 49 > > <https://reviews.apache.org/r/20465/diff/2/?file=562356#file562356line49> > > > > Would be better to actually look in zookeeper for when the monitor > > registers itself using a Watcher than just guessing that 10s is long enough. > > Bill Havanki wrote: > Quite true, I'll try that. I added a loop checking for the monitor registration, instead of just the single attempt after 10 seconds. > On April 21, 2014, 5:08 p.m., Josh Elser wrote: > > test/src/test/java/org/apache/accumulo/test/functional/MonitorLoggingIT.java, > > line 57 > > <https://reviews.apache.org/r/20465/diff/2/?file=562356#file562356line57> > > > > You can easily create a monitor entry by trying to scan a table with a > > SKVI of "java.lang.String" or something of the sort. It would be good to > > ensure that you generated a message before checking. > > Bill Havanki wrote: > I'll give that a shot. Definitely better than just assuming messages will > appear. Thanks for all the quality feedback. :) This worked. Good deal. - Bill ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/20465/#review40942 ----------------------------------------------------------- On April 22, 2014, 10:30 a.m., Bill Havanki wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/20465/ > ----------------------------------------------------------- > > (Updated April 22, 2014, 10:30 a.m.) > > > Review request for accumulo and Josh Elser. > > > Bugs: ACCUMULO-2343 > https://issues.apache.org/jira/browse/ACCUMULO-2343 > > > Repository: accumulo > > > Description > ------- > > AsyncSocketAppender is a Log4J AsyncAppender with its own internal > SocketAppender. Configuration for either appender can be set on the > AsyncSocketAppender itself. An AsyncSocketAppender can be configured using a > Log4J properties file, while an ordinary AsyncAppender cannot. > > > Diffs > ----- > > core/src/main/java/org/apache/accumulo/core/util/AsyncSocketAppender.java > PRE-CREATION > > core/src/test/java/org/apache/accumulo/core/util/AsyncSocketAppenderTest.java > PRE-CREATION > > test/src/test/java/org/apache/accumulo/test/functional/ConfigurableMacIT.java > d9bed7f > > test/src/test/java/org/apache/accumulo/test/functional/MonitorLoggingIT.java > PRE-CREATION > test/src/test/resources/conf/generic_logger.xml PRE-CREATION > test/src/test/resources/conf/monitor_logger.xml PRE-CREATION > > Diff: https://reviews.apache.org/r/20465/diff/ > > > Testing > ------- > > Unit test created and passed. Also used main() method to send log messages to > a running Log4J SimpleSocketServer instance. > > > Thanks, > > Bill Havanki > >
