In the first quarter of 2021, Zoom reported that its gross margin increased to 73.9% from 69.4% in the previous quarter, driven primarily by the optimization of public cloud resources. And Zoom is certainly not the only company to recognize the importance of cloud infrastructure optimization. As companies move their workloads to the cloud and build cloud applications, they are beginning to realize that cloud over-provisioning and cloud sprawl are not just urban legends.
For startups, the cloud is an important technology because of its unprecedented scalability support. But the cloud can quickly turn into a struggle with rising costs. Here is what a16z writes in a recent analysis:
“ The top 50 public software companies currently using cloud infrastructure are losing about $100 billion in market value due to the impact of the cloud on margins – compared to managing the infrastructure itself “
How can companies cope with the long-term financial implications of using the cloud? Optimizing cloud costs is the best answer
Three companies that optimized their cloud spending
1) Spotify Developed Cost Sharing Tool To Save Millions of Dollars
Service ownership is a key issue in lowering cloud costs. Keeping track of teams’ accountability for cloud computing and lowering costs is a pain. To deal with this, Spotify has developed its own solution called Cost Insights, which tracks the company’s cloud computing spending. In this way, Spotify allows engineers to take charge of cloud spending.
Spotify also helps its developers by offering optimization strategies such as portal autoscaling solutions. An internal crowdsourced document called Our Cookbook allows engineers to share their thoughts on what worked for them in terms of system optimization to help other teams.
“ Soon, developers began treating cost optimization as a game – showing off their victories and motivating other teams to play as well. Spotify plans to add leaderboard serviceability to play on these social and serious expense control segments“
What’s the bottom line? Spotify has cut its yearly cloud spending by a huge number of dollars – all by assisting engineers with settling on more brilliant asset distribution choices.
2) Segment optimized infrastructure and increased margin by 20%
The Segment was able to reduce infrastructure costs by 30%, all within 6 months, despite a 25% increase in traffic volume. how? It’s all thanks to the step-by-step optimization of infrastructure decisions.
An example is shown below. The Segment has an internal validation service written in Node.js that validates incoming messages to ensure they meet the message format. The group ran the service in a containerized format with one full vCPU and 4GB of memory allocated to each. To publish 200,000 messages per second, each container processed 250 messages per second, so the segment needed 800 of these containers. The service continues to get the same quota, but the team has rewritten and optimized the logic.
Coincidentally, Segment is no odder to cloud cost enhancement. The team shared another great story on their blog about how they handled the rapidly growing cloud costs and reduced AWS billing by $ 1 million annually.
“In the cloud era, costs can increase tenfold overnight as a result of sudden volume growth or one-row configuration changes. Companies need to make quick adjustments to stay profitable.
Here’s how you’ve improved your gross profit by 20% in 90 days.
— Segment (@segment) December 29, 2020”
3) La Fourche migrated to an alternate VM and saved 69.9%
Online grocery store La Fourche launched its containers on Amazon Elastic Kubernetes Service (EKS), and soon the cloud bill grew from $ 1,000 to $ 10,000. That’s when the company’s CTO decided that optimization couldn’t wait. If La Fourche had waited longer, he could be the next victim of an endless cycle of rising cloud bills and long-term savings plans.
La Fourche started by analyzing cloud invoices in detail. The organization went to the CASTAI Savings Report and ran the specialist in read-only mode on the EKS foundation. Investment funds reports show that moving to another virtual machine can essentially diminish costs. Previously, LaFourche used 15 t3.2xlarge instances and 2 t3.xlarge instances. At run time, the analysis incurred a cost of $ 4,349.95.
“ Moving these workloads to five c5a.2xlarge instances will result in a significant reduction of 69.9% ($ 1,310.40) instead. Next month, LaFourche received a bill that fell $ 3,000 “
Cloud cost optimization is a simple achievement here. You can stay in the cloud, but the cost is halved.