LoRaWAN Academy - AWS stack error

Application Development AWS
14

First off, I’m super new to all of this, so when you ask me, “Am I overlooking anything?” I can only say, “I have no idea. You tell me.” :wink:

But the version you linked to certainly seems correct, so my question is: Is there something I need to do on AWS to set up a beanstalk? From the video, it appears that all of that is supposed to be taken care of for me, but I’m not sure. I’ve followed the quick start to the letter and I keep getting this error.

At this point, I have to assume there’s something wrong with my AWS setup. I’ll see if one of my students can figure anything out. If there is any other information that might be helpful, let me know.

15

Can you try another region?

16

I tried US-west (Oregon) and got the same result.

Last night, I was able to use EB to start an application based on Go, completely independently of TTN.

Is anyone else having this issue? Or can anyone who’s following along confirm that they can make a new stack?

17

Hi Johan,
Is this issue resolved yet? I still have error message. Any work around for this?
Thank you

18

@datth22 Does this mean you had the same problem? I don’t know if that’s good (it’s not just me!) or bad (there is something deeper going on that’s going to take some time to fix), but @johan let me know what I can do to help test things.

I’ve also opened a forum post on AWS: https://forums.aws.amazon.com/message.jspa?messageID=892174 to see if anyone there has any ideas.

19

Hi SJKear,

What was the fix for this issue as I still have this error? Can you share what solution name/ platform you are running atm so I can adjust the template accordingly?

Thank you.
DAt

20

I dived in this and this is an error in the AWS documentation.

Where the documentation says:

64bit Amazon Linux 2018.03 v2.10.0 running Go 1.11.5

This should be:

64bit Amazon Linux 2018.03 v2.10.1 running Go 1.11.5

This has been fixed in our CloudFormation template – a build is running and will be deployed automatically.

1 Like
21

Hi @datth22. With my anomaly the ‘Go solution stack’ version number in TTN’s ‘cloudformation.template’ needed updating - which @johan kindly did.

Looking @johan latest post, it again looks like a version numbering issue and the TTN’s ‘cloudformation.template’ template has been updated so hopefully this should resolve your anomaly. It’s bizarre AWS has implemented it in this way…

22

I can confirm it works for me. Danke!

23

A related (?) AWS stack error … starting from the [AWS Quick Start instructions]((https://www.thethingsnetwork.org/docs/applications/aws/quick-start.html), I get an error when trying to create a new stack by clicking on (e.g.) US us-west-2. The error is

Template validation error: Template format error: YAML not well-formed. (line 308, column 13)

The same error occurs for the other regions.

Tad

24

FWIW, looking at the template file, maybe it’s just an extra tab?

The last two lines (308 and 309) might be wrongly indented:

   - Namespace: "aws:elasticbeanstalk:application:environment"
      OptionName: "TTN_SHADOW_DELTA_FPORT"
      Value:
        Ref: ThingShadowDeltaFPort
    - Namespace: "aws:elasticbeanstalk:application:environment"
      OptionName: "AWS_THING_TYPE"
      Value:
        Ref: ThingTypeName
    - Namespace: "aws:elasticbeanstalk:application:environment"
      OptionName: "AWS_CLOUDWATCH"
        Value:
          Ref: CloudWatch
25

Thanks for reporting. I fixed this just now

New build is running, please wait ~10 minutes on CD

26

Hello Johan,

I have had a similar problem as the others, and could reach to CREATE_COMPLETE without many issues by replacing the SolutionStackName item with the newest supported by AWS, i.e. SolutionStackName: “64bit Amazon Linux 2018.03 v2.11.0 running Go 1.12.1”.
The main issue now, is that I cannot see any sync Things from TTN. Message says at the aws console: " You don’t have any things yet. A thing is the representation of a device in the cloud."
I already tried to re-create the stack with a shorter ThingSyncInterval (1m, etc.), tried several other choices, and even wait for about 1/2 day. It does not seem to work.
What else could a try?
Thx!

27

Hello @guillermo_pallarolas

Can you follow those instruction, https://www.thethingsnetwork.org/docs/applications/aws/troubleshooting.html and post any error you can find ?
This is likely to be an access problem either on the TTN or AWS bound. Check your Application settings for the integration and that the integration is allowed to IoT Core via the policy.

One thing you can do to test it out is to see if there is any traffic incoming in IoT Core. If that is the case, then check, the App Access key. It should have the “devices” right for the sync to work.

Cheers

28

Hi Sylvain, thanks for your quick reply!
Yesterday I deleted the whole thing, and tomorrow will start again. I will post then the error messages according to your link. Just for the record, the only thing that was not working is the synchronization and things shadow. We could pretty much pass messages from TTN to AWS and later to DynamoDB. This is Ok for now, but we will most probably create a custom app and use the TTN SDK to access devices config from it. I find the whole AWS config, debugging and troubleshooting very complicated.
Thanks for your support!
Guillermo

29

Then something is wrong with the AppKey or the AWS Policy.

Thank you for your feedback, I feel the same way for the debugging but its AWS being AWS. We will definitely try to improve that with the V3 integration.

Cheers

31

Hi Sylvain again,
after deleting our application in TTN console, creating a new one at TTN, and creating a new stack… again we reached to the point that the devices config are not being synchronized (passing the messages to dynamo db is working pretty well though). I followed the instructions within the link you sent earlier, and from the logs the only thing that caught our attention are the following lines (the rest seems to be fine):
From /var/log/app-1.log

WARN Failed to set device remotely DevID=nodouno error=InvalidRequestException: Can not create new thing with depreated thing type:lorawan
status code: 400, request id: 96266b94-5a4e-11e9-bfa9-555efab6ca35
WARN Failed to set device remotely DevID=nododos error=InvalidRequestException: Can not create new thing with depreated thing type:lorawan
status code: 400, request id: 97ebc2b2-5a4e-11e9-af32-0358dee54a39
WARN Failed to set device remotely DevID=nodouno error=InvalidRequestException: Can not create new thing with depreated thing type:lorawan
status code: 400, request id: 97f25210-5a4e-11e9-b7f1-0bab508b41a9
WARN Failed to set device remotely DevID=nododos error=InvalidRequestException: Can not create new thing with depreated thing type:lorawan
status code: 400, request id: 99b55f82-5a4e-11e9-b8a3-a92e605c5d30
WARN Failed to set device remotely DevID=nodouno error=InvalidRequestException:

32

Looks you the ThingType has been deprecated on your account. This might be due to the integration failing during clean up correctly or a manual action.

Go to Manage > Types > lorawan > Action (right - top) > Undeprecate.

Hope it helps.

Best

33

Dear Johan,

I might be wrong but I’m failing to execute the AWS IoT integration as I’m experiencing a similar problem to others; the version of Go referenced in the CloudFormation template is possibly not current on AWS and it fails with the following message:

“No Solution Stack named ‘64bit Amazon Linux 2018.03 v2.11.0 running Go 1.12.1’ found. (Service: AWSElasticBeanstalk; Status Code: 400; Error Code: InvalidParameterValue; Request ID: d7”

AWS indicate “64bit Amazon Linux 2018.03 v2.11.1 running Go 1.12.4” as current.

Am I correct is understanding this to be a cause of problems?

with my thanks,

Dominic.

34

Yes, same old problem. I updated the solution stack again, please wait for CD to pass; ~10 minutes.