How to Reduce Ballooning AWS Costs

Integrating your business into cloud providers such as AWS and Azure is considered a modern way to have a serverless environment and lose your existing servers. However, this may turn out to be more expensive than running them on a server environment. Over time, you may experience monthly costs gradually creeping upwards with no easy ways to cost save. Main culprits in a serverless setup are:

• Lambda Code Inefficiencies
• Unoptimized Long-Term Storage
• Overkill Database Setup for handling a few peaks in demand

Getting to grips with cost analysis can help with the obvious blunders, but in-depth analysis can be tricky. Sometimes a simple tweak of code can save hundreds of monthly costs. Experienced similar problems in your company? Fill out the survey below.

Case Study: RDS SQL Server Costs

One example of where we have helped a client save significant amounts of money is on their RDS SQL server costs. Each month, snapshots were being taken of their database for accounting and backup purposes. Over several years, these snapshots ended up increasing costs and nearly doubling their RDS bill, costing them just as much as the servers themselves! To sum up:

• Not designed as a long-term storage medium
• AWS charges for each manual snapshot made
• Snapshots include unused reserved space for expansion

Our solution? Automated Backup Strategy!

By writing a series of Lambda functions, we created an automated backup strategy to backup the relevant database in the instance, straight to an S3 bucket. From here, a trigger to a Lambda can be applied to compress the data even further, and a policy on the S3 bucket to move the backups to a long-term storage tier such as Glacier storage. Result?

• 99% Cost Saving
• Data size reduced to between 50-100Gb
• Glacier storage costing just $1 for the same 5 snapshots per month
• Instant saving and paid for itself well within the first month