Investigate issue in autoscaling where the wrong node has been removed

[sanderegg]

A weird case happened today.
Autocaling might have removed the wrong node from the nodes list.
A node with IP 10.0.3.129 came in at around 7:08
Then another node with IP 10.0.3.12 came in at around 7:15
Later at 13:00:

• computer with IP 10.0.3.129 is still in AWS but not in the nodes list
• computer with IP 10.0.3.12 is not in AWS and was removed (probably by the autoscaling app), but is still in the nodes list as Down and Drain

Investigate whether some wrong handling could have gone up in autoscaling service

Graylog logs: https://monitoring.osparc.io/graylog/search/64777e171bbbd13c0565e1e5?q=%2210.0.3.12%22+OR+%2210.0.3.129%22&rangetype=relative&from=28800
CloudTrail: https://us-east-1.console.aws.amazon.com/cloudtrail/home?region=us-east-1#/events?ResourceName=i-03e858b7431e66ced

[sanderegg]

issue confirmed yesteday:
computer with IP 10.0.2.15 shown as Down Drained, while machine with IP 10.0.2.155 is running and not part of the swarm