create cluster fails with AddressLimitExceeded

[jmor...@redhat.com]

I run the installer and got this error:

./openshift-install create cluster

? SSH Public Key /Users/jmorales/.ssh/id_rsa.pub

? Base Domain openshiftlabs.io

? Cluster Name test

? Pull Secret {"auths":{"cloud.openshift.com":{"auth":"xxxx,"email":"xxxx"}}}

? Platform aws

? Region us-east-1

INFO Creating cluster...

ERROR

ERROR Error: Error applying plan:

ERROR

ERROR 1 error occurred:

ERROR * module.vpc.aws_eip.nat_eip[1]: 1 error occurred:

ERROR * aws_eip.nat_eip.1: Error creating EIP: AddressLimitExceeded: The maximum number of addresses has been reached.

ERROR status code: 400, request id: 1276cd57-19cd-49ab-b2ff-0308e877037f

ERROR

ERROR

ERROR

ERROR

ERROR

ERROR Terraform does not automatically rollback in the face of errors.

ERROR Instead, your Terraform state file has been partially updated with

ERROR any resources that successfully completed. Please address the error

ERROR above and apply again to incrementally change your infrastructure.

ERROR

ERROR

FATAL Error executing openshift-install: failed to fetch Cluster: failed to generate asset "Cluster": failed to create cluster: failed to apply using terraform

[ment...@redhat.com]

The default number of EIP for a single AWS Region is 5 and the installer is trying to create 6. You need to modify your AWS account to allow these 6 EIP to be created.

Mak

[jmor...@redhat.com]

Is that something I can do myself?

[jmor...@redhat.com]

Also, this seems like a bad limitation to anyone trying OpenShift and being a noob. I didn't find any description of this limitation/pre-requirement anywhere.

[jmor...@redhat.com]

Reply to myself. Requested limits increase to amazon, by looking at EC2 cloud console, limits for EIPs and request increase. 

[ment...@redhat.com]

Yeah, I agree that make things more difficult while using AWS for the first time. This limitation is there[1]. 

Elastic IP Addresses (IPv4)

Resource	Default limit	Comments
Elastic IP addresses per region	5	This is the limit for the number of Elastic IP addresses for use in EC2-VPC. For Elastic IP addresses for use in EC2-Classic, see Amazon EC2 Limits in the Amazon Web Services General Reference.

[1] https://docs.aws.amazon.com/vpc/latest/userguide/amazon-vpc-limits.html

[ads.gare...@gmail.com]

After installing on N. Virginia on Friday without problem I I encountered this issue today. It seems the region had 5 AZ's up on Friday but 6 today!

Is there any way to change the install config to limit the number of AZs used? I tried to install to another US region with 3 AZs and encountered other errors  that appear to be coming from Docker and related to self-signed certificates

[wk...@redhat.com]

On Monday, December 17, 2018 at 9:23:44 AM UTC-8, ads.gare...@gmail.com wrote:

Is there any way to change the install config to limit the number of AZs used?

You can set 'zones' [1] in your machine pools in install-config (e.g. in defaultMachinePlatform [2]).  YAML excerpt would be something like:

platform:

  aws:

    defaultMachinePlatform:

      zones:

      - us-east-1a

      - us-east-1b

I haven't tested that though.  And until [3] and some follow-up work lands, the Terraform configuration that's creating all the NAT gateways, etc. doesn't know which zones you'll be using and creates subnets and NAT gateways, etc. in all of them.  So currently there's not much to do in us-east-1 and other regions with lots of zones except sitting tight and waiting for those fixes to come down the pipe :/.

[1]: https://godoc.org/github.com/openshift/installer/pkg/types/aws#MachinePool

[2]: https://godoc.org/github.com/openshift/installer/pkg/types/aws#Platform

[3]: https://github.com/openshift/installer/pull/792

[tom.g...@nginx.com]

A bit of a tangent but if you select phone contact they call you immediately and will put in the request over the phone (still takes 15 minutes to propagate).

If you select web it'll take a while