OK，I ran into another problem. The master can't get started, only worker nodes are running. The following error is received after launching the cluster.
boto.exception.EC2ResponseError: EC2ResponseError: 400 Bad Request <?xml version="1.0" encoding="UTF-8"?> <Response><Errors><Error><Code>InvalidParameterCombination</Code><Message>Non-Windows instances with a virtualization type of 'hvm' are currently not supported for this instance type.</Message></Error></Errors><RequestID>60aa8aa0-acf3-4bc8-b2b4-120f0b7d1e36</RequestID></Response> On Fri, Jan 10, 2014 at 1:40 PM, Chen Jin <[email protected]> wrote: > Thanks, the patch works. > > On Fri, Jan 10, 2014 at 12:48 PM, Shivaram Venkataraman > <[email protected]> wrote: >> The EC2 scripts do support HVM instances -- Its just the i2.2xlarge >> ones were recently added by Amazon and we haven't added the mapping to >> handle them in our scripts. Should be fixed when >> https://github.com/apache/incubator-spark/pull/385 is committed. >> >> On Fri, Jan 10, 2014 at 11:51 AM, Ankur Chauhan <[email protected]> >> wrote: >>> Hi, >>> >>> I ran into the same problem, the AMI that is used but he EC2 scripts is a >>> paravirtualization one. For i2 and c3 nodes you need the newer HVM >>> virtualization. I think the ec2-scripts don't support HVM instances. >>> I ended up manually setting spark or you could try using the chef recipe to >>> do it too (search in github). >>> >>> -- Ankur
