Hello. We are running Cloud Foundry v215 deployed to vSphere and are looking to implement drs_rules to improve fault tolerence per this document https://bosh.io/docs/vm-anti-affinity.html. The current YAML structure of our resource pools looks like this (generated using spiff):

resource_pools:
- cloud_properties:
cpu: 2
disk: 32768
ram: 16384
env:
bosh:
password: REDACTED
name: runner_z1
network: cf1
stemcell:
name: bosh-vsphere-esxi-ubuntu-trusty-go_agent
version: latest

But the YAML structure in the referenced document looks like this:

resource_pools:
- name: runner_z1
cloud_properties:
datacenters:
- name: my-dc
clusters:
- my-vsphere-cluster:
drs_rules:
- name: separate-hadoop-datanodes-rule
type: separate_vms


I am having trouble restructuring the YAML implementing drs_rules and getting a successful deployment. BOSH accepts the format but the deployment fails. Can someone point me in the right direction?

Looks like my YAML format was broken. Here is the YAML configs in Github:

https://github.com/coryjett/Cloud-Foundry-DRS/blob/master/README.md

What you want is something like this (combination of two):

resource_pools:
- name: runner_z1
network: cf1
stemcell:
name: bosh-vsphere-esxi-ubuntu-trusty-go_agent
version: latest
cloud_properties:
cpu: 2
disk: 32768
ram: 16384
datacenters:
- name: my-dc
clusters:
- my-vsphere-cluster:
drs_rules:
- name: separate-hadoop-datanodes-rule
type: separate_vms
env:
bosh:
password: REDACTED

Thanks Dmitriy. Would it be possible to put this in an attachment so I can see how you have things spaced out?

We made a little more progress on this and have been able to get a successful deployment with the drs_rules section in our manifest, but the rules are not being applied on the VMWare side. Here is our current resource_pool configuration:

https://raw.githubusercontent.com/coryjett/Cloud-Foundry-DRS/master/RPConfig.MD

```
- cloud_properties:
cpu: 2
disk: 32768
ram: 16384
datacenters:
- name: DC2
clusters:
- 'PaaS POC': {resource_pool: Resources}
drs_rules:
- name: separate-dea-nodes-rule
type: separate_vms
env:
bosh:
password: REDACTED
name: runner_z2
network: cf2
stemcell:
name: bosh-vsphere-esxi-ubuntu-trusty-go_agent
version: latest
```

FYI...we have since upgraded to v217 as well.

We have submitted a pull request to resolve this issue. Seems the vSphere manifest generation is lagging a bit. https://github.com/cloudfoundry/cf-release/pull/795