Monday, April 10, 2017

Finding all the AWS resources in our account

Recently I was talking to someone about creating an account with AWS. The immediate response I got was, `AWS was charging me continuously month to month although I was not consuming any of the service`. Many times I noticed that the users of AWS start a service and forget to stop the same, the same happened to me a couple of time. Once, I started an ELB (Elastic Load Balancer) and forgot to stop the same, the ELB was running the whole night. Luckily, the hourly rate for the ELB was less and I was not charged much for the same.

One way to identify the problem is to create a Billing Alarm as shown in the previous blog.  Once we are notified of the billing, we need to identify the AWS resources which we are using and then take an action on them. But, there are 13 Regions and many services within each Region. It's difficult to go through all the services in each Region, to figure out what are all the AWS resources which we are consuming. One way is to use `Resource Groups`.

1. Go to the AWS Management Console and click on the `Resource Groups` and the `Tag Editor`.

2. In the Regions Box, select all the Regions and in the `Resource types` select `All Resource Types` and then click on `Find Resources`. All the Resources which are being used from this account will be shown as below.

3. Now, we can click on the link in the `Go` column of a particular resource and take some action like stopping it, if not required.

One of the main disadvantage of the Cloud is that the resources can be spawned in a few seconds, without much consideration if is actually required or not. Let's say the EC2 instance type is not meeting our Non Functional Requirements (NFR), then another instance can be easily started without fine tuning the existing EC2. Slowly, we end up with a lot of unnecessary AWS resources.

Also, note that not all the AWS resources will be charged. For example, a Security Group is not charged, so I can create a Security Group and forget about it. While, I cannot do the same thing with an EC2 instance.

No comments:

Post a Comment