[StarCluster] Launch failure in 0.95.4 w/no spot and different master/compute set

Justin Riley jtriley at MIT.EDU
Thu Apr 24 10:45:46 EDT 2014


John,

This is a regression from the changes in 0.95.4. I've committed a fix
for this to github already. I will release 0.95.5 that adds r3 instance
type support and fixes this issue. Thanks for reporting.

~Justin

On Thu, Apr 24, 2014 at 02:11:10AM +0000, Lilley, John F. wrote:
>    Hi All,
>    I’m having a new problem with Starcluster 0.95.4 while launching clusters
>    without spot enabled *and* the master instance is a different size from
>    the execution host (in this case the master is m3.medium and the execution
>    hosts are set to m3.2xlarge) *and* with only when cluster size =1, master
>    only. I have already successfully launched three spot enabled clusters up
>    with the exact same configuration in 0.95.4 but now need to create another
>    cluster that will be on-demand only for users submitting jobs that must
>    not be interrupted by fluctuations in the spot market. 
>    I’ve pasted the output during failure below. If I simply enable spot the
>    error does not occur *or* if I set cluster size from 1 to 2 (which adds a
>    single compute/execution instance in addition to the master) the problem
>    does not occur. I’ve created a completely fresh configuration to make sure
>    it wasn’t some weird formatting issue.
>    Thanks,
>    John
>    >>> Using default cluster template: ondemand
>    >>> Validating cluster template settings...
>    >>> Cluster template settings are valid
>    >>> Starting cluster...
>    >>> Launching a 1-node cluster...
>    >>> Creating security group @sc-new5...
>    Reservation:r-97d0929e
>    !!! ERROR - InvalidParameterValue: Minimum instance count must be greater
>    than zero
>    Traceback (most recent call last):
>      File "/usr/lib/python2.7/site-packages/starcluster/cli.py", line 274, in
>    main
>        sc.execute(args)
>      File "/usr/lib/python2.7/site-packages/starcluster/commands/start.py",
>    line 244, in execute
>        validate_running=validate_running)
>      File "/usr/lib/python2.7/site-packages/starcluster/cluster.py", line
>    1626, in start
>        return self._start(create=create, create_only=create_only)
>      File "<string>", line 2, in _start
>      File "/usr/lib/python2.7/site-packages/starcluster/utils.py", line 112,
>    in wrap_f
>        res = func(*arg, **kargs)
>      File "/usr/lib/python2.7/site-packages/starcluster/cluster.py", line
>    1641, in _start
>        self.create_cluster()
>      File "/usr/lib/python2.7/site-packages/starcluster/cluster.py", line
>    1163, in create_cluster
>        self._create_flat_rate_cluster()
>      File "/usr/lib/python2.7/site-packages/starcluster/cluster.py", line
>    1196, in _create_flat_rate_cluster
>        force_flat=True)
>      File "/usr/lib/python2.7/site-packages/starcluster/cluster.py", line
>    966, in create_nodes
>        resvs.append(self.ec2.request_instances(image_id, **kwargs))
>      File "/usr/lib/python2.7/site-packages/starcluster/awsutils.py", line
>    523, in request_instances
>        **shared_kwargs)
>      File "/usr/lib/python2.7/site-packages/starcluster/awsutils.py", line
>    624, in run_instances
>        return self.conn.run_instances(image_id, **kwargs)
>      File "/usr/lib/python2.7/site-packages/boto/ec2/connection.py", line
>    935, in run_instances
>        verb='POST')
>      File "/usr/lib/python2.7/site-packages/boto/connection.py", line 1177,
>    in get_object
>        raise self.ResponseError(response.status, response.reason, body)
>    EC2ResponseError: EC2ResponseError: 400 Bad Request
>    <?xml version="1.0" encoding="UTF-8"?>
>    <Response><Errors><Error><Code>InvalidParameterValue</Code><Message>Minimum
>    instance count must be greater
>    than zero</Message></Error></Errors><RequestID>08116944-4d1c-4e14-96c7-cdd485a06152</RequestID></Response>

> _______________________________________________
> StarCluster mailing list
> StarCluster at mit.edu
> http://mailman.mit.edu/mailman/listinfo/starcluster

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
Url : http://mailman.mit.edu/pipermail/starcluster/attachments/20140424/d6695542/attachment.bin


More information about the StarCluster mailing list