An energized Five Talent team just returned from the AWS re:Invent 2015 conference in Las Vegas armed with a cadre of new tools leveraging Amazon services. The company also just swept all five AWS certifications, making it one of Amazon’s (approximately) 1,300 top tier Consulting Partners. Leading the effort from Five Talent are CTO Ryan Comingdeer and AWS Engineer Lorenzo Aiello. The two share their thoughts on Amazon’s clear shift towards DevOps, its new services, and what it all means for their clients.
RYAN: Before the conference, I wasn’t expecting to see such a focus on DevOps. Amazon has really shifted its development efforts and new product focus from IT support to developer support. Now that they have a solid infrastructure, it’s clear they’re building more platform as a service solutions. They’re solving for platform specific pain points, standardizing best practices and giving us tools that vastly improve time to market. It’s the first time I feel like they’re really impacting our software development team’s efficiency.
In a lot of ways, AWS is still a young company taking an agile approach to problem solving. They’ve only been around for nine years and are up against companies like Microsoft Azure and Google. The biggest difference I see between AWS and the competition is that they are building solutions from scratch. In comparison, Microsoft typically buys something and then tweaks it to fit with their environment.
Amazon’s philosophy is that everything breaks, all the time. As a result, developers know they need to create their stuff so that it’s redundant and scalable and assume that services will fail. And when it does, AWS offers backup solutions to fix it. This approach means they write things from the bottom up. It also means that the way that developers use their tools is consistent across the board. It’s a more integrated system.
Five Talent has been working with AWS for a while beta testing new products and services. I think that engaging developers early on has definitely helped them produce some game-changing solutions when it comes to cloud computing. When you distill it, the benefits for both developers and clients are:
1. Rapid deployment to the Cloud
2. Lower costs
3. Greater security
4. Global redundancies
5. Deeper business intelligence
1. Rapid Deployment to the Cloud
AWS gives developers powerful tools for deploying to the Cloud faster than ever. With innovative solutions focused on mobile app development, IoT, and data migration, AWS services ensure projects get off the ground quickly with higher success.
Mobile Hub is a perfect example of how AWS has changed DevOps. Normally the first 15% of any project is building the framework. With AWS, we can build off of something that has already been tested and proven, leveraging Amazon’s already scalable platform. It’s basically a wizard for developing mobile apps. We can build on the fly without having to reinvent the wheel every time, which means we get apps off the ground much faster.
Internet of Things (IoT) Platform
The same is true for AWS IoT (Internet of Things), a communications platform that includes both the local firmware and actuators in the Cloud. Developers have been working on the framework for IoT for a couple of years because it’s really taking off, but the AWS IoT platform takes it a step further. They’re providing standardized best practices and more consistent tools for rapid deployment to the Cloud. In the end, that equals greater stability, security, scalability and speed to market.
LORENZO: Another key product for rapid deployment is Amazon’s API Gateway. One of the benefits of using the gateway is that it supports lambda functions, which allows us to get rid of servers altogether. Lambda allows us to upload code to a completely server-less environment. This means the team doesn’t have to spend time maintaining a server-based infrastructure. Instead, we simply upload snippets of code that are executed by the gateway. Lambda is executed on the fly and will scale infinitely.
RYAN: With Firehose, AWS has successfully solved some significant big data challenges. We’ve had several discussions with potential clients who will need this. There’s a current Amazon product out there called Kinesis, which is basically a funnel for streams. You can pipe large amounts of data and Kinesis stores it just long enough for you start processing it on the back end. But you have to write that processor.
AWS realized that everyone in the world was writing their own processors, so they built an alternative processor called Firehose that facilitates the storage of your inbound data. This is another part of Amazon’s shift from just supporting infrastructure to solving real market pains. They are looking at what developers are doing over and over again and addressing those issues directly.
RDS Migration Tool
LORENZO: Migrating data to the Cloud or from the Cloud can be a big issue for companies, especially for large enterprises. Historically, migrations involved downtimes for cutovers or migrations lasted months or years to avoid downtime. In particular, proprietary database systems like Oracle often required re-engineering to make it work properly in the Cloud. In response to this, Amazon created a step-by-step migration tool design to aid in quickly migrating critical databases from legacy systems into the AWS Cloud.
Amazon also created a piece of hardware called Snowball that’s designed to transfer petabytes of data in a couple days, and it’s only the size of a desktop computer. Amazon ships you their device, you plug it into your local infrastructure and transfer gigabytes of data per second. Once the data is copied, FedEx comes to pick it up and a couple of days later you have petabytes of data in the Cloud. In the past, you might invest an entire team and months to do a multi-petabyte migration. Now, a small group can do it in a week.
2. Lower Costs
Whether you’re a startup or an enterprise company, AWS services deliver large computing capacity with scalability, security, and high-powered business intelligence – at extremely affordable prices. The good news is that as economies of scale increase, the cost to use AWS will continue to drop.
RYAN: Without a doubt, AWS services are getting cheaper all the time. Their average price reduction is between 13-65% year over year (read more about AWS price reductions). It’s a race to the bottom, and clients are reaping the benefits because you only pay for what you consume.
LORENZO: EC2 spot instances are another great example of how AWS is driving down costs. By selling their unused capacity at significantly lower prices, AWS maximizes their capacity utilization while customers reap the benefits of EC2 instances magnitudes cheaper than standard pricing. With the recent introduction of Spot Fleets as well, a variety of spot instance types in multiple zones can be requested and scheduled just like on-demand instances.
3. Greater Security
As a proven, reliable platform, Amazon protects clients against security vulnerabilities and common attacks by offering a globally available, redundant and secure application. AWS is trusted by closely regulated industries to high security level organizations around the world.
RYAN: For clients of any size, AWS is more secure than having a computer behind locked doors. Or even in a data center. Amazon has passed HIPPA, PCI, and they host NSA, CIA and all organizations with high security level requirements. They have best practices for each of those security concerns as far as HIPPA, PCI and dealing with financial information. They’ve earned their trusted reputation and they continue to improve their capabilities
AWS services such as the API Gateway also prevent DDoS types of attacks or hacking attempts. At Five Talent, we write APIs all the time for our projects and we have a documentation tool that we use. Those APIs can be used by things we write, but also service third party companies too so they can access our APIs. With the AWS API Gateway, we’re still writing them ourselves, but we’re putting a funnel in front of our APIs so we can more easily monitor traffic coming in. The gateway allows us to focus on and secure traffic coming through.
Web Application Firewall (WAF)
LORENZO: The new AWS Web Application Firewall (WAF) has a similar function to the gateway when it comes to security, and applies to any size company. WAF adds another layer of protection for web applications to prevent common web exploits or vulnerabilities such as SQL injection or cross-site scripting in addition to having the ability to configure custom attack patterns that may affect application-specific vulnerabilities.
WAF is designed to work in conjunction with CloudFront (AWS’s Content Delivery Network) to provide a globally available, redundant and secure application that can resist attacks that may affect availability, security, or excessive resource utilization. Because WAF is a managed global service, any new attack vectors can be quickly mitigated by customizing security rules and attack patterns on the fly. Like the rest of the AWS platform, WAF is affordable and scalable because clients are only paying for rules and requests.
4. Global Redundancies
The AWS global infrastructure is resilient and high performing, at a level their competitors can’t touch. From number of regions to points of presence, Amazon’s approach to design for failure makes global outages a thing of the past.
As Ryan mentioned, AWS knows that things break, and they plan for it with multiple layers of redundancies and recommended best practices for when things go awry. Using AWS, you can quickly create a globally resilient infrastructure that just can’t compare to the competitors.
AWS has 11 regions worldwide with 136 points of presence while Google has only 4 regions and 70 points of presence. This means that applications are almost three times more resilient to major outages or natural disasters with AWS. Additionally, boasting almost twice as many points of presence, AWS can deliver and cache content from more locations local to the end user resulting in faster performance and more satisfied customers.
5. Deeper Business Intelligence
AWS products like Quicksite offer developers powerful reporting tools that search Amazon data repositories and return valuable business insights – at significantly lower costs.
RYAN: For me, the most exciting AWS product at the conference was Quicksite (read on making sense of customer data in AWS repositories). On almost all of our enterprise level software we come down to extensive reporting at the very end of the project. We have created report builders in the past and have also tried a lot of third party business intelligence tools. For a client, these tools typically range from $75,000 to $300,000 per year. Quicksite is competing with those tools, and is naturally built into all the other services Amazon offers. The service allows you to search data in Amazon repositories like RedShift and Elastic MapReduce. And their pricing model is the same across the board – if you don’t use it, you don’t get charged for it.
Quicksite is in preview right now. However, it’s something I want to evaluate heavily for almost all of our projects going forward as a reporting tool we can build into our enterprise solutions.
The Bottom Line on Certification
For Five Talent whatever service we’re using from AWS, the AWS certification means that we can handle a professional level of DevOps and Solutions Architect work. We know how to architect the infrastructure and how to how to write the code to best use the AWS platform. When we’re developing on the AWS platform, we know the accuracy of our estimates, we’ve done it before, it is scalable, and most importantly, it is secure.
We’ve got new products coming in the door every day. Between Mobile Hub, the API Gateway, QuickSite and a few others, we now need to get the team together and do some training so we can leverage AWS services for our clients and ourselves. Those certifications aren’t easy, but the return on that investment is already paying off for Five Talent. We’re excited to be a part of what Amazon is doing, and where they’re heading next.