Horotonworks Data Cloud error autoscaling token expired

[#1]

I am provisioning a cluster using version 1.14.0 of Hortonworks Data Cloud. The Autoscaling status slider is in the off position, but cluster provisioning fails with the following error on the auto scaling token:

Infrastructure creation failed. Reason: com.amazonaws.AmazonServiceException: The security token included in the request is expired (Service: AmazonAutoScaling; Status Code: 403; Error Code: ExpiredToken; Request ID: cf2aebb0-2cec-11e7-8985-4100dda4669c)

The autoscaling tab on the failed cluster in the cloud controller, shows the cluster trying to autoscale to 3 nodes:

[#2]

Hello,

So you are using Controller 1.14.0. A few more questions:

- What HDP version are you attempting to launch?

- What Cluster Type?

- What is your Worker + Compute count? Spot instances? Etc?

- Maybe just post the .json (that would be used for the cluster create CLI) and we can see all the options you selected.

Thanks!

Jeff

[#3]

I tried HDP 2.6 on us-east, hive+spark2.1, 3 workers, pretty much standard everything

[#4]

The error happens when you use default settings, just specify cluster name and Ambari password.

[#5]

I think there is a bug where it is checking the spot bidding prices when you don't even request spot I tried resizing my cluster twice, both fails hdp26resizefail2.png

I have a feeling if I wait until after 5 or 6pm it will work

[#6]

What worked for me last time is creating the cloud controller in a region different than us-east.

You could also try using newer version 1.14.1: http://docs.hortonworks.com/HDPDocuments/HDCloudAWS/HDCloudAWS-1.14.1/index.html

[#7 · Accepted solution]

In case this helps, back in March I got this error when using HDP 2.5 EDW-ETL in us-east.

In my case, for some reason the infrastructure creation took too long (1 hour instead of 5 minutes) and the security token expired during that time.

In addition to what Jeff said, it may help to check the CloudFormation UI on the AWS console, find your stack and post the data from the “Events” and “Resources” tabs.

[#8]

This happened twice yesterday. The third try in the afternoon worked (all AWS us-east) Tried to add one worker today and it failed again, same error. Something is really broken here hdp26secondfail.pnghdp26addworkerfail.png

[#9]

@jeff any ideas on this?

[#10]

@cdub

@DL

What worked for me is creating the cloud controller in a region different than us-east. When I use us-east with this specific cloud controller version, I seem to get various errors.

You could also try using the newer cloud controller version 1.14.1 with us-east: http://docs.hortonworks.com/HDPDocuments/HDCloudAWS/HDCloudAWS-1.14.1/index.html