This is basis of the demo that Circonus was using at HashiConf NAPA 2016 and continues to be improved

It is based based on kelseyhightower's awesome Hashconf EU 2016 presentation. Well, more than based on, a minor update to include telemetry data for each of the components (Consul, Nomad, Vault, as well as Fabio).

Completely awesome example showing the power of Hashicorp's products, with a little bit of Circonus magic thrown in.

So, if you follow the instructions below, you'll end up with a deployment that looks something like this:

Overview

Once you have completed the setup, you will have:

a 3 node server cluster (ns-1, ns-2, ns-3) running Consul, Nomad, and Vault servers. Each member of the cluster will AUTOMATICALLY create a corresponding set of Circonus metrics for each of the servers (Consul, Nomad, and Vault). These metrics will include the all important latency metrics (represented via histograms in Circonus), as well as basic information on CPU, Memory, and Disk utilization
5 client machines with Nomad configured in agent mode, and, again, each of these machines will AUTOMATICALLY create a set of Circonus metrics with similar information as that of the Servers.
A MySQL instance (to store the VAULT secrets)
A load balancer (to give you one public IP)

Now that you have the basic infrastructure, you'll then launch two nomad jobs

Consul in agent mode
Fabio, a sophisticated software load balancer

These nomad jobs are configured to run a single allocation on each of the 5 clients, and each will produce a corresponding set of Circonus metrics

Finally, you will launch a simple application called 'hashiapp' via one of the server cluster machines (ns-1), and you can experiment with starting and stopping an application, adjusting the allocation count (there are versions with 3, and 10 allocationssd) and again, viewing the

Prerequisites

Bootstrap the HashiStack on Google Compute Engine [Once this is done you'll see your metrics flowing into your Circonus account.]

gcloud compute ssh ns-1

git clone https://github.com/vynjo/hashiconf-napa-2016

cd hashiconf-napa-2016

Create the Hashiapp Policy and Token

export VAULT_ADDR=http://ns-1:8200

vault policy-write hashiapp vault/hashiapp-policy.hcl

vault token-create \
  -policy="hashiapp" \
  -display-name="hashiapp"

Edit jobs/hashiapp.nomad job with TOKEN AND MYSQL address

env {
  VAULT_TOKEN = "HASHIAPP_TOKEN"
  VAULT_ADDR = "http://vault.service.consul:8200"
  HASHIAPP_DB_HOST = "CLOUD_SQL:3306"
}

Create the Hashiapp Secret

vault read mysql/creds/hashiapp

vault write secret/hashiapp jwtsecret=secret

Service Discovery with Consul

nomad plan jobs/consul.nomad

nomad run jobs/consul.nomad

nomad status consul

consul join nc-1 nc-2 nc-3 nc-4 nc-5

consul members

Start Load Balancing with Fabio

Edit the jobs/fabio.nomad file and replace CIRCONUS_API_TOKEN with your Circonus api token

nomad run jobs/fabio.nomad

nomad status fabio

Hashiapp Job

Submit the hashiapp service job.

nomad run jobs/hashiapp-v1-c3.nomad

nomad status hashiapp

Viewing Logs

nomad fs -job hashiapp alloc/logs/hashiapp.stderr.0
nomad fs -job hashiapp alloc/logs/hashiapp.stdout.0

Send Traffic

curl -H "Host: hashiapp.com" http://<loadbalancer-ip>:9999/version

while true; do curl -H "Host: hashiapp.com" http://<loadbalancer-ip>:9999/version; sleep 1; done

Scaling Up

Run the jobs/nomad run jobs/hashiapp-v1-c10.nomad which starts 10 allocations instead of 3

nomad run jobs/hashiapp-v1-c10.nomad

Then you can switch to v2 by running

nomad run jobs/hashiapp-v2-c10.nomad

Rolling Upgrades

Change up the jobs by running jobs/hashiapp-v1-c3.nomad or jobs/hashiapp-v2-c3.nomad

or via your local machine:

gcloud compute ssh ns-1 nomad run hashiconf-napa-2016/jobs/hashiapp-v2-c10.nomad

vynjo / hashiconf-demo