I'm trying to trace down why the ZookeeperInstance defaulting to localhost. Is the code you pasted all your testing on the client, or are you doing other things in there which may be related? Additionally, I'm going to have to ask you to update your code since you seem to be a few weeks behind on the version of trunk your running. It may have gotten resolved by now.
John On Wed, Jul 11, 2012 at 10:09 PM, David Medinets <[email protected]>wrote: > I have an accumulo single-node system running at Linode. I can see the > Accumulo monitor at http://accumulo.affy.com:50095. It all looks good. > But I want to develop Accumulo applications on my Windows laptop. I > blissfully whipped out some Java code: > > String instanceName = "development"; > String zooKeepers = "zookeeper.affy.com"; > String user = "root"; > byte[] pass = "NOT_THAT_STUPID".getBytes(); > > ZooKeeperInstance instance = new ZooKeeperInstance(instanceName, > zooKeepers); > instance.getConnector(user, pass); > > When I run this program, I see: > > 12/07/11 22:00:14 WARN conf.SiteConfiguration: accumulo-site.xml not > found on classpath > .. > 12/07/11 22:03:31 INFO zookeeper...., connectString=localhost:2181 > .. > watcher=org.apache.accumulo.core.zookeeper.ZooSession$AccumuloWatcher@91e321 > > Should Accumulo be looking at the ZooKeeperInstance constructor for > the Zookeeper connection information? I'd rather not need the > accumulo-site.xml file on my Windows laptop. That seems rather bad > practice to require a copy on every developer's workstation. > Especially because it contains passwords. >
