Ive been using the standard hadoop monitoring tools with:
ssh -L 9100:localhost:9100 -L 9101:localhost:9101 -o ServerAliveInterval=10 -o StrictHostKeyChecking=no -i key.pem hadoop@ec2-blah-blah-.compute-1.amazonaws.com
And then just using my browser to navigate to localhost:9100 to monitor the instances. I've been installing ganglia though, and want to know how to access the ganglia web interface via the same method.
I read that the default port for ganglia is 8649, so I tried:
ssh -L 8649:localhost:8649 -o ServerAliveInterval=10 -o StrictHostKeyChecking=no -i key.pem hadoop@ec2-blah-blah-.compute-1.amazonaws.com
and going to localhost:8649, but that didn't work.
I can use ganglia using the method described by amazon's emr documentation. But this web tunneling method should work too, right?