Blog
Testimonial
Jaret Chiles
Hey thank you everybody for joining us again. This is another Mission talk on Dr. Werner Vogels’ CTO Keynote. This is Jaret Chiles, Vice President of Consulting Services, and I have Tim with me here today.
Tim Banks
Hi everybody. TIm Banks, Consultant here at Mission.
Jaret Chiles
We are excited. We’re going to have another one of these videos for you. We’re going to talk a little bit about our reactions and some of the topics that were covered today which was really cool. There were a lot of good themes around reinvention which has been of course what we have seen throughout all of these sessions. If you’re interested, please go to the Mission blog and you can look back and see some of the previous recordings and we’ll probably have more coming of these as well. But, yeah, Tim and I are just going to go back and forth and share some of our thoughts on this.
I will say, you know, one thing, Tim, that I thought was really cool was just how he was able to tie the reinvention theme into how companies were surviving and being resilient and even serving clients in different ways through the pandemic. That was one of the first big things that came out and he shifted from there into some of the classic Werner MO around cloud best practices and what AWS is doing to make those kinds of things easier for developers and operations teams. Let’s talk about some of the pandemic implications. What did you see that was interesting to you that came out of the session today?
TIm Banks
Well I think first and foremost, I think AWS has been really beating the invention drum. I really think that when we consider what we thought 2020 was going to be like on January 1, 2020 versus what it is like now on December 15, 2020, like that reinvention has been so necessary. We’ve seen our own customers along with many others have to completely change how they do their business: completely change how they have their infrastructure, completely change how they have their software deployment, change their business models altogether, try to be agile in adjusting these things. So talking about reinvention this year and how businesses really do it I think is super important and super appropriate considering all the things that people had to do. There has been so much innovation with people trying to reinvent themselves in order to stay afloat here in 2020.
One of the things I thought they brought up, they brought up a company, Ava, which was a women’s reproductive health company that gathered biometric data to help women make good data-driven decisions on their reproductive healthcare. But in the time of COVID, they took that biometric data and now they’re using it to detect early warning signs when someone has COVID. It’s interesting because they said classify themselves as a data company, not a healthcare or reproductive company. And so, what do you do with that data? I thought it was really awesome having that.
Jaret Chiles
Yeah. It’s really cool seeing that. When you have these type of cloud practices in place where if you’re able to pivot quick into it, there were a few companies they talked about that did that, not only did it help make some of these companies more resilient but it gave them that opportunity to pivot quickly with this infrastructure behind them or these options behind them to serve clients differently and I thought like early detection out of that platform. They already have the sensors available, they had access to data, all they had to do was go relearn some algorithms and things like that in order to help identify those. It sounds like they’re right on the cusp now going through some final testing stages where it is actually really a viable solution.
Tim Banks
Yeah. Like 20,000 people or something like that. I think it’s interesting because, you know, we look at some of the other changes that companies had to make because of the currencies with the pandemic whether they’re working from home or whether they’re, not just working from home but where their kids are and what the infrastructure that children need in order to have daycare or get educated. Some of the practices that you have to get around because now instead of having maybe one person, if you were working from home before, one person using up all the bandwidth, now you got all these people: online classes, working from home, some kids are streaming video because they are too young for classes. That gets all kinds of spikey traffic. How many people did not use Zoom several hours a day before this year are doing it now?
Jaret Chiles
They said 30 times growth. 30 x the growth that they say when the pandemic kicked off. If you just think of a classic architecture in your own data center, no one anticipates 30 x growth, right. So I think that’s really incredible and the ability to have that scale and pivot strategy is critical to their success. There were other examples, not even necessarily pandemic oriented but similar dealing with the spikey traffic. It’s a classic cloud use case. They talk about Lego. Lego is a great example where they talked about these massive spikes in certain times of the year when they launch new products. I was particularly interested in the Millennium Falcon launch; they talked about that.
Tim Banks
Of course. Well I think it’s one of the things that are funny. I think that it is very topical with the pandemic because people are home now. People are home with their kids. Disney is releasing all of these Star War things, so yeah, you’re going to buy the Millennium Falcon when it comes out and everyone is going to jump to try to buy it at once because what else are you doing?
Jaret Chiles
And for them, it’s about focusing on their differentiator, right? They’re a product company developing these awesome Lego things and being creative and running a data center they said themselves is not their core differentiator. That’s not what they needed to be focused on. So not only did this give them the scalability they needed but it helped them focus too.
Tim Banks
Well I think it’s interesting to talk about. One of the things he talked about was running your own data center versus running things in cloud and what the overall impact on the environment is; the social environmental impact about running data. So we talk about AWS is going to be completely either carbon neutral - not carbon neutral, sorry, but completely sustainable over the next, I think, it said few years?
Jaret Chiles
They invest so tremendously in their data centers efficiencies, just out of a matter of necessity and things like that. It’s really interesting when you tie that back into thinking of all the aspects here, you can actually see a direct correlation to things like optimizing your architecture in your applications for, what looks like cost optimization, but in the end what you’re actually really doing too by adopting their platform and driving optimization in is an environmental responsibility: the ability to reduce your impact.
Tim Banks
Yeah. It’s optimizing. When you think about some of the best practices when it comes to using containers, right? We call in bin packing, you’re playing Tetris. You’re trying to get as many of the containers on that one box as you can reasonably have and still maintain resiliency. But when you have that kind of density in there, it operates more fuel efficiently. Instead of spinning up 30 of 40 servers to run a workload, it can spin up 3 or 4. That’s huge. Or spin down stuff that you’re not using and then you can be more choosy about what you spin up. Again, to beat a dead horse when it comes to the armbase processors, they run faster, they run cooler, and they run less expensively; less power consumption for the same performance. So, you can cost optimize by using an arm processor, you can power optimize by using an arm processor, and you still get more computing power out of that for most of your workloads.
Jaret Chiles
Definitely making some recommendations along those lines and callouts. You know, there are a lot of things that are going to run just fine on arm as is but there are some things that will not. But the upfront investment in development is actually miniscule compared to the long term operational impact and cost efficiencies and, you could even say, environmental impacts by going and making some of these investments. He was definitely a stronger promoter of that today.
Tim Banks
Yeah. I think it's important. Again, when we talk about reinvention, it’s changing the way you do business or maybe changing the way that you have been writing software or that you have been deploying things onto something that's going to be a little bit newer and more sustainable. Not just from a business standpoint, from an environmental standpoint. Even the process of considering that is how businesses can make a difference and maybe change the way you did it before. If you never considered the environmental impact of your software development practices or your infrastructure practices, maybe it’s about time you do.
Jaret Chiles
Yeah.
Tim Banks
I think one of the other real big things he talked about was dependability.
Jaret Chiles
Definitely.
Tim Banks
That was a huge, huge portion of it. I think when we talk about going on the cloud in the first place, one of the main advantages folks having by being in a public cloud is to say “hey, I want to be able to have multiple instances, I want to be able to have multiple availability zones, I want to be able to pull here and here and here so I can have something be more dependable.”
One that he really dove into - what dependability is. Dependability is a business decision. What is a critical application? What is critical infrastructure? How does your business define that? What can you say is an acceptable amount of down time for that application or that infrastructure for you to say that it’s dependable? And then how do you shoot for that? I think one of the things that he really talked about in order to define that is observability. You have to be able to have insight into your stack, into your infrastructure, into your software deployment, into your security practices, into all of these little different facts and aspects of your infrastructure in order to figure out if you’re doing the right things, if you have that uptime, if you have the service expectations that you’re setting for your customers.
Jaret Chiles
They really went into it this whole session but to me when he got really into dependability, and as you mentioned all the different aspects of dependability he talked about security and all these things, to me this was like classic cloud best practices. What’s really cool about it is all the investments that they’re making into making cloud best practices easier. Just because we’ve said cloud best practices for years now doesn’t mean it’s always been easy to do it well. Things like the VPC reachability analyzer, the IAM analyzer, all these tools that they’re rolling out there to actually help with security. I mean, let’s talk about chaos engineering, that was a big topic.
Tim Banks
That’s one that’s near and dear to my heart as an operations and SRE focused person. Chaos engineering is something that has been making a big push in the industry, especially in the DevOps and SRE world. Seeing AWS now launch that as a whole service with the fault simulator I think is fantastic. It really gives orgs a good chance to “hey let’s see what happens when we throw a wrench in here” for real because a lot of times before “hey what is your fault tolerance look like?” “oh we’re going to take down this thing in a very predictable manner because we already know how to fix it”, right? But when the pager goes off at 3 o’clock in the morning, it’s never that.
Jaret Chiles
No. Never!
Tim Banks
So it gives you a chance to really see what can happen and how to architect around that.
Jaret Chiles
Nice shoutout to the classic Chaos Monkey and Netflix for helping pioneer the way, in some ways, of Chaos Engineering. But, now making it accessible and readily available to everybody coming soon in 2021. I’m really excited about that. But, the thing that I also liked that he called out - it’s not just about identifying the solution, the technical gaps in the environment, these simulations actually help them build really important operational muscles that you need to learn to develop to respond to incidents. It's the classic: everything fails, all the time. You should assume for everything. You can’t normally just plan for everything very well, but when you start to make things programic through tools like the AWS Fault Simulator, that makes it a lot more of a reality.
Tim Banks
So, the thing is that resilience is not a technical solution. Resilience is a cultural solution. Your work culture has resilience. And when your work culture or your org or whatever it is has resilience as a value, it comes out in your infrastructure, it comes out in your software deployment, it comes out in what you write. I think this enables that, is all it is.
The other thing I want to talk about that was really awesome was some of the things was the announcement of the managed services for Grafana and Prometheus because in order to know what dependability is, you have to be able to get the metrics on it. What you measure is what you’re going to do. So, a lot of people have had sometimes issues scaling out Prometheus / Grafana; scaling out their observability to meet the size of their infrastructure and the shortcut is always to let some stuff go. Well we can’t measure these things with everything else so we’re going to measure these things. Without that, you don’t have real observability: you don’t know how dependable you are, you don’t know if you’re actually resilient. So, having this as a managed service so that you can scale out some of these end points and see the data and have it be readily available and have it be in an observable state. If you've never used Grafana before, it is such an amazing tool for visualizing some of these time series and metrics data that you have in your infrastructure. So, I’m super excited to see that. You get a lot of control without a lot of the headache. It’s fantastic.
Jaret Chiles
Yeah. I think it’s so common, especially for smaller companies who haven’t been through the hyper scale growth and things like that, to underestimate the complexibility of good observability at scale. He used examples of things like being able to just correlate time stance with infrastructure based metrics and things like that and really the importance of having application level logging - the logs matter and these are the services that really help drive more observability into that full stack through the process. Really glad to see that. That’s going to make a lot of things easier at scale. I’m sure a lot of developers are really excited about that as well and operations teams.
Tim Banks
Yeah. I think another thing that I really need to make sure that we talk about is that we talk about all the things that go into making something dependable, that go into resiliency, and it’s a lot. It’s not just that you can flip the switch and do it. Especially if you're one of the smaller companies that has not gone into hyper scale and you’re experiencing those growth pains and your technical debt is starting to rack up, that's where Mission can come in and help you out with that. We have the insight and the expertise to be able to kind of unravel some of that yarn ball and figure out what we can do, what needs to be done, maybe what needs to be reinvented, and what you can salvage. That way you can get a better grasp on what you can do, what your business is good at, and then you can build a product or build infrastructure that’s going to be lucrative for you. I think the important part about this is that when you go through the process of reinvention, it’s never usually going to be easy. If it was easy everybody would do it. But, we can help you bridge that gap whether it’s knowledge, whether it’s training, whether it’s hands on, there’s a lot of opportunities that we have to help you out on your path to reinvention. It’s always exciting.
Jaret Chiles
Yeah. Tim, I would be remiss to try to wrap a better closing than that. I think you really just summed it up well. There’s been so many amazing announcements and so many good best practices and things that have been shared throughout the session today and other sessions but it’s a lot to take in and without the right kinds of muscles and partners, it’s a really difficult thing to go and really take full advantage of and really get the full value out of. There was so much more in this session that was covered today. I think I had 200 some odd lines worth of notes and I barely caught it all myself. But, I really enjoyed it and I think this was a great one.
Definitely, everybody again, make sure you check out some other videos if you’re interested in seeing some of our other recaps. Not recaps, but reactions to some of these sessions and expect more to come. Reach out to Mission if there is any way we can help you. Tim just wrapped that up real well. We’re here for you. These types of services, these launches, make what we’re able to provide our clients even more valuable. So, I’m really excited about ramping up and getting more familiar with some of these tools and having our teams do so and being able to help everybody else take the full value of AWS and wrap into their operations.
Go Back
Share this post