Upgrade manager fails to actually upgrade nodes because of AWS token issues

[shrinandj]

Is this a BUG REPORT or FEATURE REQUEST?:
BUG REPORT

What happened:
Upgrade manager fails because it cannot get AWS credentials. It consistent fails with the following errors:

time="2019-12-03T05:09:43Z" level=info msg="retryable: RequestError: send request failed\ncaused by: Put http://169.254.169.254/latest/api/token: net/http: request canceled (Client.Timeout exceeded while awaiting headers) -- ec2metadata/GetToken, will retry after 2.898737352s"
time="2019-12-03T05:09:49Z" level=info msg="retryable: RequestError: send request failed\ncaused by: Put http://169.254.169.254/latest/api/token: net/http: request canceled (Client.Timeout exceeded while awaiting headers) -- ec2metadata/GetToken, will retry after 3.287156464s"
time="2019-12-03T05:09:51Z" level=info msg="retryable: RequestError: send request failed\ncaused by: Put http://169.254.169.254/latest/api/token: net/http: request canceled (Client.Timeout exceeded while awaiting headers) -- ec2metadata/GetToken, will retry after 3.542636997s"
time="2019-12-03T05:09:51Z" level=info msg="retryable: RequestError: send request failed\ncaused by: Put http://169.254.169.254/latest/api/token: net/http: request canceled (Client.Timeout exceeded while awaiting headers) -- ec2metadata/GetToken, will retry after 4.523316716s"
time="2019-12-03T05:09:52Z" level=info msg="retryable: RequestError: send request failed\ncaused by: Put http://169.254.169.254/latest/api/token: net/http: request canceled (Client.Timeout exceeded while awaiting headers) -- ec2metadata/GetToken, will retry after 4.209153918s"
time="2019-12-03T05:09:57Z" level=info msg="retryable: RequestError: send request failed\ncaused by: Put http://169.254.169.254/latest/api/token: net/http: request canceled (Client.Timeout exceeded while awaiting headers) -- ec2metadata/GetToken, will retry after 4.761618292s"

This is most likely because of aws/aws-sdk-go#2972.

What you expected to happen:
There should be no errors and upgrade-manager should actually perform the upgrade.

How to reproduce it (as minimally and precisely as possible):
Just update the launch-config of an IG and run upgrade-manager.

Anything else we need to know?:
Manually changing the version of aws-sdk to v1.25.0 (something < 1.25.38) works.

Environment:

• rolling-upgrade-controller version
• Kubernetes version :

$ kubectl version -o yaml

Other debugging information (if applicable):

• RollingUpgrade status:

$ kubectl describe rollingupgrade <rollingupgrade-name>

• controller logs:

$ kubectl logs <rolling-upgrade-controller pod>