There are someways to createCloudwatchDashboards such as creating them manually by selecting widgets from AWS Console, with Cloudformation etc.
I’ve decided to create them with Python because in DevOps literature, there is no such a thing as manually creating something. I also didn’t want to use Cloudformation because I like scripting and we have manyapplicationstomonitor in our company, thus, I needed something to iterate over our environments and create dashboards for eachofthem.
To deploy this configuration file and make the Cloudwatch Agent read it which we installed via Ansible by this post, you need to be sure that you installed Cloudwatch Agent to your instance properly. After installing the agent, you need to put this configuration file under /opt/aws/amazon-cloudwatch-agent/etc directory.
When you start the agent after putting your configuration file, Cloudwatch service will automatically turn it into a .toml file under the same directory and reads that file.
If your agent refuse to start when you run this:
Make sure you don’t have any syntax errors in metric configuration or common-config.toml where you put your proxies.
This configuration file will only retrieve metric data(CPU Usage, Ram Usage, Free Storage, Network Usage etc.). Which is really, REALLY useful when you want to monitor your instances so that you can see if something goes wrong. You can even initialise alarms to warn you when something goes over the threshold which I am going to investigate in another post.
In this post, I will be explaining how to install Cloudwatch Agent using an Ansible Controller Host machine which is used for initializing and provisioning other machines.
You will need:
Basic knowledge of Ansible
A Controller machine to run the playbook remotely
A Controller machine with Ansible installed
A Remote machine with SSH Daemon and Python interpreter installed
A Remote machine with proper Policy attached (Allowance for PutMetricData)
Make sure you have a proper configuration .json file, check here
Using a Controller EC2 instance makes it easier to manage other instances since in this way, you don’t need to install Ansible each machine you manage because it works with SSH Daemon and an up-to-date Python interpreter.