Awesome! Would be good if this could be checked during start and communicated appropriately though. Room for improvement :)
On Jan 27, 2012, at 12:20 PM, Tim Robertson wrote: > Thanks Akash and Andrei for your help today, > > I tried from home and it all just worked. Must be some DNS thing in > the office - I'll try and get to the bottom of that next week. > > Great stuff to see how easy it is with Whirr. Let the HBase benchmarking > begin! > > Tim > > > On Fri, Jan 27, 2012 at 5:26 PM, Akash Ashok <[email protected]> wrote: >> Hi Tim, >> Could you please do this small test. >> >> Extract Whirr onto an EC2 instance itself . Try to spawn an HBase cluster >> from there. My guess is that it would spawn it flawlessly. If so then its a >> classic case of DNS resolution . >> >> Cheers, >> Akash A >> >> >> On Fri, Jan 27, 2012 at 9:49 PM, Tim Robertson <[email protected]> >> wrote: >>> >>> Sure thing. I've gone through the 30k lines of log file, and don't >>> see anything other than the fail to connects: >>> >>> 2012-01-27 14:28:23,406 DEBUG [jclouds.compute] (user thread 11) << >>> tailerr(0) >>> 2012-01-27 14:28:23,407 INFO >>> [org.apache.whirr.actions.ScriptBasedClusterAction] (pool-4-thread-3) >>> configure phase script run completed on: us-east-1/i-096dab6c >>> 2012-01-27 14:29:25,729 DEBUG [jclouds.compute] (user thread 1) >> >>> running >>> [./configure-zookeeper_hadoop-namenode_hadoop-jobtracker_hbase-master >>> tail] as [email protected] >>> 2012-01-27 14:29:26,186 DEBUG [jclouds.compute] (user thread 1) << tail(0) >>> 2012-01-27 14:29:26,186 DEBUG [jclouds.compute] (user thread 1) >> >>> running >>> [./configure-zookeeper_hadoop-namenode_hadoop-jobtracker_hbase-master >>> tailerr] as [email protected] >>> 2012-01-27 14:29:26,638 DEBUG [jclouds.compute] (user thread 1) << >>> tailerr(0) >>> 2012-01-27 14:29:26,639 INFO >>> [org.apache.whirr.actions.ScriptBasedClusterAction] (pool-4-thread-4) >>> configure phase script run completed on: us-east-1/i-116dab74 >>> 2012-01-27 14:29:26,639 INFO >>> [org.apache.whirr.actions.ScriptBasedClusterAction] (main) >>> Successfully executed configure script: [output=No directory, logging >>> in with HOME=/ >>> >>> No directory, logging in with HOME=/ >>> No directory, logging in with HOME=/ >>> No directory, logging in with HOME=/ >>> starting jobtracker, logging to >>> /var/log/hadoop/logs/hadoop-hadoop-jobtracker-ip-10-79-37-92.out >>> Warning: $HADOOP_HOME is deprecated. >>> No directory, logging in with HOME=/ >>> starting master, logging to >>> /var/log/hbase/logs/hbase-hadoop-master-ip-10-79-37-92.out >>> No directory, logging in with HOME=/ >>> , error=12/01/27 13:29:12 INFO ipc.Client: Retrying connect to server: >>> 107.20.125.231/107.20.125.231:8020. Already tried 1 time(s). >>> >>> My properties (stripping comments): >>> >>> whirr.cluster-name=hbase >>> whirr.instance-templates=1 >>> zookeeper+hadoop-namenode+hadoop-jobtracker+hbase-master,1 >>> hadoop-datanode+hadoop-tasktracker+hbase-regionserver >>> whirr.provider=aws-ec2 >>> whirr.identity=${env:AWS_ACCESS_KEY_ID} >>> whirr.credential=${env:AWS_SECRET_ACCESS_KEY} >>> whirr.hardware-id=c1.xlarge >>> whirr.image-id=us-east-1/ami-da0cf8b3 >>> whirr.location-id=us-east-1 >>> >>> Thanks, >>> Tim >>> - online in a couple hours >>> >>> >>> On Fri, Jan 27, 2012 at 5:13 PM, Andrei Savu <[email protected]> >>> wrote: >>>> On Fri, Jan 27, 2012 at 6:07 PM, Tim Robertson >>>> <[email protected]> wrote: >>>>> Reverse DNS seems to work if I understand you correctly: >>>>> >>>>> $ host 174.129.76.5 >>>>> 5.76.129.174.in-addr.arpa domain name pointer >>>>> ec2-174-129-76-5.compute-1.amazonaws.com. >>>> >>>> Looks fine to me. There is something else wrong. Can you share the >>>> .properties file? >> >>
