CODEWITHSUNDEEP
amazon-web-servicesamazon-ec2cloud-foundry
owaism
owaism

Reputation: 1188

Cloud Foundry router cannot find api.xx.xxxx.com/info (AWS)

Finally managed to successfully deploy cloud foundry to AWS. Mostly following instructions from http://docs.cloudfoundry.org/deploying/ec2/bootstrap-aws-vpc.html

Its failing at the validation step that is to get a success response for the following:

curl api.subdomain.domain/info

Of course I have substituted the subdomain and domain appropriately.

I am getting the error: 

404 Not Found: Requested route ('api.XX.XXXXX.com') does not exist.

The request is coming till the Cloud foundry router router_z1. And I can see this error in the logs for router_z1.

Here is output of my bosh vms command:

------------------------------------+---------+---------------+--------------+
| Job/index                          | State   | Resource Pool | IPs          |
+------------------------------------+---------+---------------+--------------+
| unknown/unknown                    | running | medium_z1     | 10.10.16.254 |
| unknown/unknown                    | running | medium_z2     | 10.10.81.4   |
| unknown/unknown                    | running | small_errand  | 10.10.17.1   |
| unknown/unknown                    | running | small_errand  | 10.10.17.0   |
| api_worker_z1/0                    | running | small_z1      | 10.10.17.20  |
| api_z1/0                           | running | large_z1      | 10.10.17.18  |
| clock_global/0                     | running | medium_z1     | 10.10.17.19  |
| etcd_z1/0                          | running | medium_z1     | 10.10.16.20  |
| hm9000_z1/0                        | running | medium_z1     | 10.10.17.21  |
| loggregator_trafficcontroller_z1/0 | running | small_z1      | 10.10.16.34  |
| loggregator_z1/0                   | running | medium_z1     | 10.10.16.31  |
| login_z1/0                         | running | medium_z1     | 10.10.17.17  |
| nats_z1/0                          | running | medium_z1     | 10.10.16.11  |
| router_z1/0                        | running | router_z1     | 10.10.16.15  |
| runner_z1/0                        | running | runner_z1     | 10.10.17.22  |
| stats_z1/0                         | running | small_z1      | 10.10.17.15  |
| uaa_z1/0                           | running | medium_z1     | 10.10.17.16  |
+------------------------------------+---------+---------------+--------------+

The only change that I made in the CF deployment manifest was to eliminate instance from zone 2. The reason being AWS default limit for number of instances on EC2 in a particular region is 20.

Any pointers on how to resolve this issue will be appreciated.

Upvotes: 1

Views: 1776

Answers (1)

owaism
owaism

Reputation: 1188

Figured out the problem. Couple of issues:

  1. In the CF deployment manifest make sure the system domain property is <BOSH_VPC_SUBDOMAIN>.<BOSH_VPC_DOMAIN>. That is if you have reserved cf.example.com for cloud foundry PaaS. Make sure cf.example.com is what system_domain property in your cloud foundry deployment manifest refers to. Infact example.com should not appear in your deployment manifest anywhere without cf.. Through out the deployment manifest it is always cf.example.com
  2. Do not use '@' in any of the passwords within the deployment manifest. I have logged a bug for this in cf-releases: https://github.com/cloudfoundry/cf-release/issues/527

Upvotes: 1

Related Questions