brooklyn-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Aled Sage <aled.s...@gmail.com>
Subject Re: AWS Fails to start
Date Thu, 16 Apr 2015 17:09:55 GMT
Andrea,

Yes, that makes sense to get Brooklyn working again in this scenario.

Would be good to also fix in jclouds (I'm guessing the next minor 
release for 1.9.1 will be in approx 6 weeks?).

Aled


On 16/04/2015 11:59, Andrea Turli wrote:
> Please find here the jira issues:
>
> https://issues.apache.org/jira/browse/BROOKLYN-140
> https://issues.apache.org/jira/browse/JCLOUDS-889
>
> Looking closer at the documentation at
> http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/virtualization_types.html
> it seems that `All current generation instance types support HVM AMIs.`
> Would it make sense to give a better score to the AMIs that are explicitly
> supporting hvm?
>
> Andrea
>
> On 16 April 2015 at 11:41, Aled Sage <aled.sage@gmail.com> wrote:
>
>> +1
>>
>> Can you raise a jira issue on Apache Brooklyn, and also one on jclouds
>> (cross-referencing)?
>>
>> In Brooklyn, we set a jclouds `TemplateBuilder.imageChooser()`, but that
>> is just given a list of valid image choices to choose from. So the problem
>> I presume is in the jclouds code that generates the list of valid images.
>>
>> Aled
>>
>>
>>
>> On 16/04/2015 11:24, Alex Heneveld wrote:
>>
>>> Andrea/All-
>>>
>>> But we already ask jclouds to filter the images against the hardware --
>>> so why is it even considering this incompatible imageId?
>>>
>>> Best
>>> Alex
>>>
>>>
>>> On 16/04/2015 09:09, Andrea Turli wrote:
>>>
>>>> Duncan,
>>>>
>>>> I tried your yaml and it is not working for me as well.
>>>> I think the problem is related to this error message that you should see
>>>> in
>>>> the summary tab of your `Empty large system`
>>>>
>>>> Failure running task invoking start[locations] on 1 node (gfHA5Sr8):
>>>> Error
>>>> invoking start at EmptySoftwareProcessImpl{id=CHjSrqrz}:
>>>> org.jclouds.aws.AWSResponseException: request POST
>>>> https://ec2.eu-west-1.amazonaws.com/ HTTP/1.1 failed with code 400,
>>>> error:
>>>> AWSError{requestId='c72a6249-31b7-44c4-a30b-013527ebe249',
>>>> requestToken='null', code='InvalidParameterCombination',
>>>> message='Virtualization type 'hvm' is required for instances of type
>>>> 'r3.large'.', context='{Response=, Errors=}'}
>>>>
>>>> It seems to be an extra-requirement enforced by ec2 when you have to use
>>>> a
>>>> particular instance type (see
>>>> http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/
>>>> virtualization_types.html
>>>> for more details)
>>>>
>>>> It should be probably enough to add a particular imageId that explicitly
>>>> supports `hvm` like ami-df6a8b9b
>>>>
>>>> Best,
>>>> Andrea
>>>>
>>>> On 16 April 2015 at 08:07, Duncan Grant <duncan.grant@cloudsoftcorp.com>
>>>> wrote:
>>>>
>>>>   I'm getting the same error when I specify minRam 4096.  Dropping the
>>>>> minram
>>>>> requirement fixes the problem (except I have a vm without enough ram).
>>>>>
>>>>> My yaml looks like:
>>>>>
>>>>> name: Empty large System
>>>>> location: jclouds:aws-ec2:eu-west-1
>>>>> services:
>>>>> - type: brooklyn.entity.basic.EmptySoftwareProcess
>>>>>     provisioning.properties:
>>>>>       minRam: 8192
>>>>>
>>>>> And from the logs the template details are:
>>>>>
>>>>> 624314 2015-04-16 13:47:27,890 DEBUG b.l.jclouds.JcloudsLocation
>>>>> [brooklyn-execmanager-eH5sU4JL-511]: jclouds found template
>>>>> {image={id=eu-west-1/ami-3fcb5348, providerId=am i-3fcb5348,
>>>>> name=RightImage_CentOS_7.0_x64_v14.2.1_EBS, location={scope=REGION,
>>>>> id=eu-west-1, description=eu-west-1, parent=aws-ec2, iso3166Codes=[IE]},
>>>>> os={family=ce       ntos, arch=paravirtual, version=7.0,
>>>>> description=411009282317/RightImage_CentOS_7.0_x64_v14.2.1_EBS,
>>>>> is64Bit=true}, description=RightImage_CentOS_7.0_x64_v14.2.1_EBS,
>>>>>    version=14.2.1_EBS, status=AVAILABLE[available], loginUser=root,
>>>>> userMetadata={owner=411009282317, rootDeviceType=ebs,
>>>>> virtualizationType=paravirtual, hypervisor=xen       }},
>>>>> hardware={id=t2.medium, providerId=t2.medium, processors=[{cores=1.0,
>>>>> speed=0.4}], ram=4096,
>>>>> supportsImage=Predicates.and(requiresRootDeviceType(ebs),Predicates.o
>>>>>
>>>>> r(isWindows(),requiresVirtualizationType(paravirtual)),Predicates.
>>>>> alwaysTrue(),Predicates.alwaysTrue())},
>>>>> location={scope=REGION, id=eu-west-1, description=eu-west-1,
>>>>>    parent=aws-ec2, iso3166Codes=[IE]}, options={userDataCksum=2f4a740b}}
>>>>> (image {id=eu-west-1/ami-3fcb5348, providerId=ami-3fcb5348,
>>>>> name=RightImage_CentOS_7.0_x64_v14.       2.1_EBS,
>>>>> location={scope=REGION,
>>>>> id=eu-west-1, description=eu-west-1, parent=aws-ec2, iso3166Codes=[IE]},
>>>>> os={family=centos, arch=paravirtual, version=7.0, descriptio
>>>>> n=411009282317/RightImage_CentOS_7.0_x64_v14.2.1_EBS, is64Bit=true},
>>>>> description=RightImage_CentOS_7.0_x64_v14.2.1_EBS, version=14.2.1_EBS,
>>>>> status=AVAILABLE[available       ], loginUser=root,
>>>>> userMetadata={owner=411009282317, rootDeviceType=ebs,
>>>>> virtualizationType=paravirtual, hypervisor=xen}}) for provisioning in
>>>>> JcloudsLocation[AWS Dub
>>>>> lin:AKIAIKSCBKOOAXIOFYRA/aws-ec2:eu-west-1@uuGaPaUU] for
>>>>> aws-ec2:eu-west-1@EmptySoftwareProcessImpl{id=AXHSzp1t}
>>>>>
>>>>> Could someone try the yaml above and see whether they have the same
>>>>> problem?
>>>>>
>>>>> Also is there anything in the log that indicates what the problem is?
>>>>>
>>>>> thanks
>>>>>
>>>>> Duncan
>>>>>
>>>>> On 10 April 2015 at 18:41, Aled Sage <aled.sage@gmail.com> wrote:
>>>>>
>>>>>   Robert,
>>>>>> There should be more info earlier in the log about the full jclouds
>>>>>> "template" being used (including instance type, AMI, etc). Could
you
>>>>>> post
>>>>>> that here please?
>>>>>>
>>>>>> If the "specified instance type" that has been auto-selected by
>>>>>> brooklyn/jclouds now requires a VPC, then perhaps others can reproduce
>>>>>> by
>>>>>> explicitly choosing the same options as were in that template.
>>>>>>
>>>>>> Aled
>>>>>>
>>>>>>
>>>>>>
>>>>>> On 10/04/2015 13:14, Alex Heneveld wrote:
>>>>>>
>>>>>>   I've confirmed AWS is working as normal for me.
>>>>>>> Robert, this suggests something with your local setup, or more
>>>>>>>
>>>>>> worryingly
>>>>>> something with your account.  Could AWS have switched a default on
some
>>>>>>> accounts so that new machines require VPC and thus add'l setup
to use
>>>>>>> them?
>>>>>>>
>>>>>>> Good to chase this down as if the latter it will affect people
getting
>>>>>>> started with Brooklyn.
>>>>>>>
>>>>>>> Best
>>>>>>> Alex
>>>>>>>
>>>>>>>
>>>>>>> On 10 April 2015 at 10:32, Robert Moss <robert.moss@cloudsoftcorp.com
>>>>>>> wrote:
>>>>>>>
>>>>>>>    Hey,
>>>>>>>
>>>>>>>> After pulling from master today, I am unable to deploy anything
to
>>>>>>>> AWS.
>>>>>>>> My
>>>>>>>> properties file is unchanged.
>>>>>>>>
>>>>>>>> Has anyone encountered this error before:
>>>>>>>>
>>>>>>>> Error invoking start at BasicApplicationImpl{id=QClWJ8Z3}:
Error
>>>>>>>> invoking
>>>>>>>> start at MySqlNodeImpl{id=SbTHpjaK}: org.jclouds.aws.
>>>>>>>> AWSResponseException:
>>>>>>>> request POST https://ec2.us-west-1.amazonaws.com/ HTTP/1.1
failed
>>>>>>>> with
>>>>>>>> code
>>>>>>>> 400, error: AWSError{requestId='efe77bfa-
>>>>>>>> b73d-40b2-82e2-8660492fb86b',
>>>>>>>> requestToken='null', code='VPCResourceNotSpecified', message='The
>>>>>>>> specified
>>>>>>>> instance type can only be used in a VPC. A subnet ID or network
>>>>>>>>
>>>>>>> interface
>>>>>> ID is required to carry out the request.', context='{Response=,
>>>>>>>> Errors=}'}
>>>>>>>>
>>>>>>>> Robert
>>>>>>>>
>>>>>>>>
>>>>>>>>


Mime
View raw message