Want To Truly Understand Your AWS Footprint? Start Here!

Apptio has an ideal starting point

We’ll be using the  Apptio Cloudability default dashboard to give an example of how to better visualize your AWS costs and usage.

• Review popular widgets that deliver immediate insights into your AWS spend
• Explain the mechanics behind the widgets and how you can adapt them to your particular needs

You’ll be looking at a multi-layered dashboard, so as we walk through this post, we’ll cover things “row by row.”

First Row: Trends

Use this top row to quickly understand your cloud spending trends from different heights.

Our first widget here, called “Monthly Estimated Spend,” is an absolute staple from our experience. This shows how the current month is progressing compared to the previous month, and how much your estimated spend (based on an average 30 day rolling window) at the end of the month will differ.

On this same row, the middle widget is a number/KPI widget that tells us how much you have spent this calendar year. This is a great starting point, but you may want to edit the widget to get a more important KPI that’s relevant to your business or team’s success. Some examples could be:

• “How many EC2 instances ran last week?”
• “What is this month’s RI coverage?”
• “How much did team X spend on RDS last week?”

The final trend widget displays resource (usage) cost per-day, including the upfront component of RIs amortized over the reporting period. We also have a trendline overlaid upon the graph to give a clear view of where your “real” spend is heading.

Having this day-to-day view, that takes into account full RI costs, without being distracted by spikey non-usage costs will establish what your clear trend is. You’ll likely find chart widgets to be the most flexible overall, with filters/dimensions/metrics to match your exact needs and the ability to overlay time series across each other (super handy for identifying correlations).

Second Row: Key Dimensions

You’ll want the means to drill into your spend by dimensions that matter to your business.

Our first widget here is another daily chart widget, over a shorter time period, but we’ve added some color by stacking with the dimension Product Name. Product Name refers to the AWS product such as EC2, RDS etc. We can use this to get a good sense of what constitutes our day to day spending, and help to make sense of any spikes. A good tip here is to swap in a business dimension which is pertinent to you, perhaps a top level tag like “business unit,” lengthen the time scale and stretch the widget.

Our second widget is a straightforward pie chart showing us the “Top 5 AWS Accounts by Spend.” Feel free to change the dimension and metric to something else that better suits your operations. Even copy the widget as to keep the original, and then create your widget from the new one.

Our final widget here is a table widget, where you can choose the exact set of dimensions and metrics to display in a tabular format. In this particular widget we are establishing what is the biggest type of cost using a synthetic Apptio Cloudability dimension called Usage Family. In that widget, we can see that EC2 and RDS instance usage are the biggest costs followed by storage. Now we have a great place to begin optimizing.

Third Row: Cost Drivers

Let’s dive into what’s truly causing fluctuations in your spend, including a diff report and list of resources that are driving your bill up.

In row three we have two table widgets. Both of those are focused on what’s driving changes to your AWS spend. The first widget is a direct comparison of last month to the month prior. The default dimensions of Account Name and Service Name are a fantastic start, pointing you exactly in the direction of your cloud spend culprit. In the above example we can see the biggest increase is $11.6k associated to EC2 in the Sandbox Account. A good idea here would be to add an additional dimension such as “business unit” or “application” to really zone in.

Our second widget is one of our most popular, enumerating your most expensive AWS resources. What are your 5-10 most expensive things? Do they include an S3 bucket, EC2 volume or Redshift instance? Identifying them quickly lets your teams investigate how to lower those costs.

Fourth row: Cloud Resource Usage

Finally, let’s go beyond cost to see actual usage trends that may be affecting your footprint.

Our first widget is a deep dive into measuring your footprint based on usage of EC2 instances. This is a chart widget, with one layer represented by columns and another a line. You should edit the widget to see how the layers are configured independently. This widget focuses on common questions our customers want answered: how many unique instances are running each day, and what are total instance hours I’m paying for.

PRO-TIP: Hover above the series to see the values for each day. Keep in mind that if an individual instance appears for just one hour in the day, it will contribute once to each the unique count and usage hours. Whereas if an instance runs all day, it will still contribute one instance to the unique instance count, but will contribute 24 hours to instance hours. Often we can identify quite a story from how the lines react to each other!

Our final dashboard widget is focused on what your data transfer costs are per AWS Product. Data transfer costs are one of our most popular items to categorise across the tool and allocate the spend to the relevant group.

Build a Cloud Cost Dashboard to Reflect Your Business

Even though we expect you’ll get fantastic insights from this dashboard we highly recommend adapting for your particular needs. A good idea could be to duplicate the dashboard and then play with with individual widgets to get the data points you want.

NOTE: All new Cloudability users will get this great new dashboard by default. If you are a current user and would like this dashboard to replace your old default dashboard do reach out to your friendly Account Services team to get going. Or start a Free Trial if you’re new!

With the right data points at hand, your teams can catch anomalies, spending spikes or underutilized resources ahead of time and make confident technical decisions to increase both savings and the efficiency of your cloud.