I wish to launch multiple Amazon Web Service EC2 instances simultaneously. For example, I hope to run 1000 models on one instance and a separate 1000 models on a second instance.

When I click Request Spot Instances I select 2 instances under the Total target capacity option. Then I click the box to the left of Maintain target capacity.

A fleet is created consisting of: m5d.12xlarge, r3.8xlarge, m5dn.8xlarge, m5n.8xlarge, r5n.8xlarge. This fleet is listed on the Spot Requests page. Capacity for that fleet says 1 of 2. Status says pending_fulfillment.

However, only one instance is listed on the Instances page. This instance is running.

Presumably I can only run the first set of 1000 models on this instance. How can I initiate the second instance so I can run both sets of 1000 models simultaneously?

Must I simply wait until the Capacity column for the fleet on the Spot Requests page says 2 of 2? If so, how long does this take? I tried waiting 25 minutes before creating this post.


I found the following error message when clicking on pending_fullfillment:

1/13/2020, 3:03:39 PM   error   spotInstanceCountLimitExceeded

I had no idea there is a Spot Instance count limit. Why would the option exist to select multiple instances if there is a limit of only one instance?

  • This question should not be closed. It is perfectly appropriate for this forum. If you feel otherwise, please explain why. – Mark Miller Jan 15 at 21:16

Is your AWS account new? There's a standard limit of 20 spot instances, but new accounts have lower limits.

It's also possible that you've hit your limit for EC2 instances - go into the EC2 console and near the top left click "limits". If you can start an EC2 on-demand instance of the same type the spot won't create, then it's the spot limit rather than the EC2 limit.

You can request a limit increase from AWS (log in before you visit this link).

Not the answer you're looking for? Browse other questions tagged or ask your own question.