Stuck on waiting for an agent

[sweetfxv]

Multiple stages from different pipelines that uses the same elastic-profile stuck on waiting for an agent.

The logs are:

[go] Received a request to create an agent for the job: [Prod-N/180/Build_Test_Push_Prod/2/Build_Test_Client_CF]
[go] No running instance(s) found to build the ECS Task to perform current job.
[go] The elastic agent profile is configured to run on a spot instance. Initiating steps to request for a spot instance.
[go] There is an open spot request matching the profile, not requesting for a new spot instance.
[go] Received a request to create an agent for the job: [Prod-N/180/Build_Test_Push_Prod/2/Build_Test_Client_CF]
[go] No running instance(s) found to build the ECS Task to perform current job.
[go] The elastic agent profile is configured to run on a spot instance. Initiating steps to request for a spot instance.
[go] There is an open spot request matching the profile, not requesting for a new spot instance.
[go] Received a request to create an agent for the job: [Prod-N/180/Build_Test_Push_Prod/2/Build_Test_Client_CF]
[go] No running instance(s) found to build the ECS Task to perform current job.
[go] The elastic agent profile is configured to run on a spot instance. Initiating steps to request for a spot instance.
[go] There is an open spot request matching the profile, not requesting for a new spot instance.
[go] Received a request to create an agent for the job: [Prod-N/180/Build_Test_Push_Prod/2/Build_Test_Client_CF]
[go] No running instance(s) found to build the ECS Task to perform current job.
[go] The elastic agent profile is configured to run on a spot instance. Initiating steps to request for a spot instance.
[go] There is an open spot request matching the profile, not requesting for a new spot instance.

GoCD version: 21.2.0
Plugin version: 7.2.1-278

[chadlwilson]

And what is happening to the spot requests on AWS - were they fulfilled? Did the instances start up and join the cluster correctly? What do the logs from the instances tell you about whether they started correctly?

Basically, this is what you will see for a number of reasons, which basically means that an instance did not start and register with the cluster for an agent task to be assigned to it, either because the spot request was never fulfilled (no capacity) or a problem starting the instance with your chosen AMI, or a problem with the chosen AMI being able to register with the cluster (could be permissions/IAM roles or networking reasons)