maryol illanit neleah

[Coleman John]

Netflix uses AWS for almost everything cloud computing. That includes online storage, a recommendation engine, video transcoding, databases, and analytics. So most of the $1 billion Netflix plans to spend on cloud services will go into Amazon Cloud Services.

These commitments may require a greater investment in AWS cloud services. Second, to satisfy shareholders and to avoid external financing for day-to-day operations, the company needs higher net profits.

To provide full cost visibility, the company deploys a custom data dashboard. The Efficiency Dashboard serves as a transparent feedback loop to its data consumers and producers. Netflix credits merging cost and usage context via dashboards for its cost-efficient architecture.

To appreciate how big of a deal that is, consider the amounts of data and different platforms Netflix needs to aggregate in one place, compute, and send to engineers so they can come up with working cloud cost optimization strategies.

The video streaming service generally uses two types of data platforms in motion and data at rest. While the first cost category involves processing transient data, data at rest systems involve physical data storage costs. Both categories include infrastructure spending.

Netflix gets its AWS billing data through the AWS Cost and Usage Report, like everyone else. You might know that the data can be tough to derive meaningful business insights from whether you consume it via S3 or CSV. It is even more challenging for decision-makers who are not data scientists.

Netflix uses a microservices architecture on AWS. Microservices architecture helps an organization to scale without additional work. It also helps maintain a cost-effective operation in the cloud and eliminates a single source of failure even if engineers change/update/upgrade multiple service areas in one go.

That agility helped the video streaming service innovate faster and cost-effectively, leading to Chaos Engineering, Spinnaker, and Global cloud, as well as the unprecedented growth Netflix sees today.

For example, you may see a high cloud bill at the end of a given month which may raise some alarms. However, you may have also signed new clients who demand more cloud usage and you may have also built a new product feature as part of a new and more lucrative client contract.

You do not need a team of data scientists or an in-house cost tool to improve your cloud costs. Instead, you can use a robust cloud cost intelligence solution to dig into your AWS bill, understand what you are spending and why, and maximize your ROI.

Netflix used their internal spot market to save 92% on video encoding costs. The story of how is told by Dave Hahn in his now annual A Day in the Life of a Netflix Engineer. Netflix first talked about their spot market in a pair of articles published in 2015: Creating Your Own EC2 Spot Market Part 1 and Part 2.

At any point in time AWS has a lot of underutilized instances. It turns out so does Netflix. To understand why creating an internal spot market helped Netflix so much, we'll first need to understand how they encode video.

Netflix gets video from production houses and studios. First, Netflix validates the source file, looking for missing frames, digital artifacts, color changes, and any other problems. If problems are found, the video is rejected.

Netflix has a huge baseline capacity of reserved instances. They autoscale in and out 10,000s of instances a day from this pool. When Netflix autoscales down they have unused capacity. Unused capacity is a waste of money and spot markets are great way of soaking up all that unused capacity while also getting important work done.

So Netflix did a genius thing, they built their own internal spot market to process the chunked encoding jobs. The Engineering Tools team built an API exposing real time unused reservations at the minute level.

Netflix started their own internal spot market for the same reason Amazon did; cloud economics are all about driving higher machine utilization. Reserving instances saves a lot of money in AWS, it makes sense to extract the most value as possible out of those instances. Every microsecond CPUs are not working is a waste of money.

Back in the day, like many people, before it became clear AWS would become the eater of all infrastructure, I was brainstorming AWS startup ideas. I was pursuing some of the ideas I later detailed in Building Super Scalable Systems: Blade Runner Meets Autonomic Computing In The Ambient Cloud.

The kicker was security. Who would run code and put their data on an a random machine without a security guarantee? This was before containers. Though I had used Jails on FreeBSD to good effect, the idea of containers never occurred to me.

My idea was something like lambda, which was why in What Google App Engine Price Changes Say About The Future Of Web Architecture, I was disappointed when GAE pivoted towards a higher granularity system:

The basis of this conjecture/vision is the development and evolution of one of GAE's most innovative and far reaching features: task queues. It's a great feature that allows applications to be decomposed into asynchronous flows. Work is queued and executed at some later time. Instead the monolithic and synchronous model used originally by GAE, an application can be completely asynchronous and can be run on any set of machines. For a while now it has been clear the monolithic front-end instances have become redundant with the fruition of task queues.

The problem is task queues are still image based. Operation are specified by a URL that terminate inside a run time instance whose code template is read from an image. An image contains all the code an application can execute. It's monolithic.

When a web client URL is invoked it executes code inside a monolithic image. It's these large images that must be managed by GAE and why Google needs to charge you more. They take resources to manage, time to initialize, and while running take memory even if your app isn't doing anything.

A different idea is to ask why can't a request terminate at a task queue work item instead? Then the monolithic image could be dropped infavor of an asynchronous coding model. Yes, GAE would have to still manage and distribute these code libraries in some fantastical way, no simple task, but this would solve the matching work to resources granularity problem that they instead solved by going the other direction, that is making images the unit of distribution and instances the unit of execution. We'll talk more about the granularity problem next.

So with this super cool task queue framework and programming model being developed I felt sure they were ready to announce that the monolithic images would disappear, instances would disappear, and there would be an even finer pay for what you use billing model as a replacement. I was wrong. Again.

Driving this upheaval is that programs run on an abstract machine that uses resources that are quantized differently than the underlying physical machines. A server comes with only so much memory and CPU. Running programs use memory even when a program is idle. Google must pay for the machine resources used by an instance. Charging only for the resources used by a program instead of all the resources used to host a program creates an unsustainable and unprofitable pricing friction between the two models.

In other words, programs are deployed in big quanta, but run in small quanta. A smaller work granularity would allow work to be schedule in idle times, which is why I think the task queue model is superior.

Depending on your subscription plan, the price hike adds up to an extra $24 or $36 you pay each year to the streaming service. Below, CNBC Select shares some ways to save on (and even benefit from) your Netflix subscription.

While you're unlikely to be happy about paying a higher monthly Netflix bill, it does mean you earn a bit more from that 6% cash back. Amex's cash back is earned in the form of Reward Dollars, which cardholders can then use as a statement credit to lower their credit card balance.

And with the U.S. Bank Cash+ Visa Signature Card, cardholders can choose to earn 5% cash back on two bonus categories each quarter, on their first $2,000 in combined eligible net purchases, then 1%. Television, internet and streaming services are counted as a bonus category and U.S. Bank's website lists Netflix as a sample qualifying merchant. Again, you can use this cash back to essentially lower your credit card bill.

T-Mobile has a "Netflix On Us" deal where qualifying cell phone plans get a free Netflix subscription. Those who aren't happy with their current cell phone provider should consider this benefit, which not only makes Netflix complimentary but also consolidates your streaming and cell phone bill.

Netflix allows you to pause your membership and come back to it. This can give you a break from the monthly subscription if you're looking to cut costs or if you're just not watching a particular show at the moment.

You just have to connect the bank account you use to pay your Netflix subscription to Experian Boost, and Experian will add your payments to your Experian credit file. Consumers can link positive payment data as far back as 24 months. Experian Boost also includes access to your FICO Score and Experian free credit monitoring that alerts you to changes on your credit report, such as new account openings in your name and balance updates.

Basic and Premium plan Netflix subscribers will now pay a little more each month for the streaming service. To help save on this cost, get a credit card that rewards streaming purchases, switch your phone plan to T-Mobile or take a pause on your subscription. And, while you're paying more for it, make sure that monthly Netflix bill is helping your credit with Experian Boost.

At CNBC Select, our mission is to provide our readers with high-quality service journalism and comprehensive consumer advice so they can make informed decisions with their money. Every personal finance article is based on rigorous reporting by our team of expert writers and editors with extensive knowledge of personal finance products. While CNBC Select earns a commission from affiliate partners on many offers and links, we create all our content without input from our commercial team or any outside third parties, and we pride ourselves on our journalistic standards and ethics.

90f70e40cf