The current trends in technology indicate that more than 60% of the businesses use cloud computing for their IT operations. Among the various cloud service providers, Amazon Web Services [AWS] is a pioneer and continues to be a leader in the cloud market.
At Flux7, we have had several magnificent encounters with clients whose needs have revolved around and within the cloud. Specifically, AWS. This post is about three such scenarios we’ve had during the last few months.
Reducing AWS spending by $5,000 per month
Moving two renewable energy giants to the cloud
Achieving high performance and low costs using AWS
Reducing AWS Spending by $5,000 per Month
Amazon Web Services delivers a rich feature set and provides services on a pay-as-you-go basis. Although this means that a customer or user pays only for the service that he/she actually used, it is quite possible to overpay and end up with a heavy monthly AWS bill.
Consider the following scenarios which depict how AWS bills could be unknowingly high, possibly due to lack of best practices.
1. Running Dev and Test instances even while not in use, say, at night, when the instances could be turned off.
2. Unused EBS volumes, private AMIs and snapshots, that consume a lot of a space, and, in turn, a lot of money.
3. Lack of autoscaling policies in place that lead to over allocation of resources, even during times of lull for your website or application.
4. Use of instance sizes that don’t fit or meet the actual requirements. Using instances larger than is needed increases cost, and using instances smaller than is needed reflects in the performance.
A similar situation was faced by our client who was overpaying by $5,000 per month. A thorough AWS audit was performed by Flux7. We concluded that with a comprehensive list of recommendations, we could reduce the client’s AWS bills up to $5,000 each month, and even more. Read the case study here to better understand more about the AWS audit and our findings .
Moving Two Renewable energy Giants to the Cloud
As mentioned before, the cloud is becoming the necessity of the day, and it’s a typical exercise of enterprises moving, or planning to move, to the cloud. Several business requirements call for cloud migration, including scalability, agility, high availability, and the list goes on. One such interesting call for migrating to the cloud was from solar panel manufacturers with applications that could be put to full use if moved to the Cloud.
Consider the following:
Solar Panel Monitoring Systems -- provide real-time data on the power consumed by a customer at any given time.
Customer Portal, Billing & Accounting System -- a comprehensive dashboard to enable customers to better understand their usage, track bills paid, and defuncts in payment.
A few other applications include billing enforcement system and firmware upgrade manager, as well as contract and quotation management software.
What benefits did we gain from the cloud-based solutions? Here are some of them:
Ability to produce higher quality software
Reduce conflicts between Devs and Ops
Create production-like Dev environments in less than five minutes
No shortage of IT staff and shorter new hire ramp-up
Improve customer engagement and satisfaction
Read the case study now to better understand how Flux7 got it done.
Achieving High Performance at Low Cost Using AWS
The third use case describes cloud usage for an online media company, Yactraq. It provided a speech2Topics API. Its service focuses on extracting data from videos for use in advertising, search and discovery. The API returns a list of topics discussed in the video at specific times. The API is also used for sentiment analysis. The API works in real-time, and customers expect the topics within minutes of submitting a video.
Yactraq’s real issues peak in times of increased demands.
Yactraq has partnered with Flux7 to handle the scaling and price optimization. At Flux7, we developed VyScale, which is a spot-strategy-as-a-service SaaS product. VyScale uses proprietary machine-learning algorithms and historical data to dynamically choose the best spot strategy.
Using AWS spot instances in our strategy led to:
60-million seconds of video successfully processed
60%-80% reduction in costs
Effective autoscaling and disaster recovery solutions
Automated solution requiring no human intervention
Our motivation to use spot instance gets credited to Yactraq’s pricing model, which is based on the number of minutes processed, and that the cost of compute directly impacts the bottom line.
As a result, we proudly earned an honorable mention at the 2013 AWS re:Invent event for our product VyScale and its use at Yactraq. Click here to learn how Flux7 was able to make such an impression .
Then read the case study about Yactraq to better understand how Flux7 improved its process.