Datadog and Cloud Concepts Flashcards
What is Datadog?
A monitoring and analytics platform. It gives IT teams visibility into the health and performance of their entire software stack.
Let me talk broadly about the problem space:
So with modern web services and applications people who use services and the businesses that rely on them really expect for things to be available and on all the time. Majority of businesses have gone through some sort of digital transoframtion and then entrusted their life-blood yah know their core business workflows are accessed and done online, and so this comes with a lot of challenges for the infrastructure and development teams who are creating and maintaining these services.
So how datadog helps these people and provides value to them is by giving them tools to monitor and observe in real-time their entire stack. So the lowest layers so that could be infrastrcutre, metrics on memory usage etc. to the top-level application code which can emit logs and measure response times, and it also can tie into all the other components such as DBs, 3rd party APIs, CDN.
So Datadog talks about offering all this via that “single pane of glass” if you will, to break down silos, and make it easy to not only react to issues and be more effective as a team and business.
What do you know about our platform and products? Talk about Datadog’s Product Portfolio.
Talk about Datadog’s competition
What is Kubernetes
What are containers?
What is the cloud? How to explain to a 7 yr old
I LOVE this question because I have a lovely little niece Moira, she’s only 4 but I picture her once she’s a little bit older and explaining this stuff to her. So I would say okay, Moira, think about your backpack that you take to nursery school, you put inside of it your stickers, your lunch, your tablet, your books, all those things, and then you carry it with you on your back, and when you need something you take it out. Your backpack is great, but it does have some limitations - you can only fit so much inside of it, you have to have it with you when you need something, and your responsible for keeping it safe. Now imagine a magical backpack, this is the cloud, your cloud pack, and it’s extremely big, 100s of times bigger then yours and it’s magical because you don’t have to carry it around with you, instead you can access it from anywhere as long as you have -an internet connection, and your special key–your password. And, the cloud takes care of keeping it safe for you. So there are many benefits to this cloud backpack, you can fit much more inside of it, you don’t have to carry it, when you need something you use your connect and your key, and it’s going to take care of storing and keeping your things safe for you.
Convince an on-prem CTO to go to the cloud. What is the value proposition of the cloud?
Okay let me take a few moments to gather my thoughts…
Write them down
So I want to go over the 6 pros I thought of and then I have about 3 cons to share and then I wanna review with you to see if I missed any
So to convince this person to go to the cloud first I would say
1. Capital Expenditure: They have a big opportunity to lower their capital expenditures. They'll no longer have to purchase, maintain and manage the lifecycle and upgrades of physical machines. There's a lot that goes into that like renting or owning space, physically securing the space, locks on the doors, card access control system, video cameras and the personelle and know-how to do that, so that's a large expense and expertise that you can pay for and offload to the cloud provider 2. Security: Second as I mentioned is security right, so not only that physical security but digital, network, and isolation, investing in pen testing, finding an open source service to manage certificate and secret rotation, all of those things you're also offloading. CSPs have whole teams dedicated to this, and yes you're going to need to follow their best practices that they have on their docs, use key vault for example, but really you can rely on them when you spin up a VM or use a serverless function that it'll be isolated the way they say it will. Now don't get me wrong here, things do happen that's a con that I'll mention later 3. Elastic: Number three is the elasticity so as your business grows you'll have to not only correctly forecast and anticipate this growth & have the machines ready to use when you need them for a spike in usage, but then you can't return them if the service scales down, when not being used they sit there idle. So another benefit is the pay for what you need usage model. 4. World-wide: And, related to this as well is the world-wide presence of CSPs, so if you're business is going international well let's set up our services in a European or Australian data center and improve the latency and response times for those users because they're not making API calls all the way to the US and back every time they wanna use us. And another aspect of multi-region is the ability to have replicas for disaster recovery. Now similar to security this is something that you have to be aware of, configure storage replicas and follow best practices, and it doesn't come for free but it's easy to set up 5. Ecosystem: And the last thing I could think of would be the ecosystem and integrations available, there are all these developer conferences, and a community, and services that integrate such as datadog right, services that tie in really well and strive to be "plug and play" rather then perhaps selecting something open source and having to configure it yourself for monitoring.
So all of those benefits combined mean that you’ll free up a lot of resources and thus can focus instead on YOUR business, your bread and butter, and add additional features for users to your product or platform. So developer efficiency and faster time to market which is good for your bottom line.
In terms of cons
1. Security and data privacy implications, so cloud providers with so many ppl using them have become a target, and if there is a breach, suppose someone was going after Bank of America and you just happened to be hosted next to them for no fault of your own, well you can impacted in that situation. 2. Lack of control if there is an outage, which is very rare, but they do happen, and you can't react you can't fix it yourself you'll have to wait for the CSP to get back online. 3. Lastly is vendor lock-in if you really get married or tightly couple everything to the CSP you selected. You're then at the mercy of their pricing and changes to services, including sunsetting or deprecating services. The 2nd team I was on in Azure, Time Series Insights I saw that it'll be sunsetted in 2025 and that they're asking folks to migrate over to Azure Data Explorer So running a business is all about risk management and it used to be that cloud was considered risky, but now not so much I would say it’s the other way, that running your own stuff is deemed more risky bc it's been proven and hardened and battle tested.
Thanks! Do you think I missed anything?
How, when, and how often do you normally
interact with customers?
Why do you want to work at Datadog?
I can name many reasons why I’d like to be a datadog, let me talk through the top 3 or so
1. Number one is the product itself and the brand, right. I want to work at a company that's a leader in what they do, and Datadog is right up there in the upper right quadrant in the magic quadrant. Customers are finding value in your products, it's solving problems & making their lives better, that really came through as I was reading customer case studies 2. Reason number two for me would be company culture - and there are multiple aspects of this, it's a tech-first company yah know a former colleague of mine is at American Express and when he tells me about some of the roadblocks he faces and lack of agility I think wow I don't want to be there, I wanna be somewhere like Datadog that was founded by developers for developers. But not only a technology first but also you're known to have a really strong product culture, known for product-led growth! I heard you all just got back from a summit! so I'm thinking wow I could thrive here, I could learn a lot here and grow my career, eventually step into a role as a manager. 3. And lastly I do wanna be somewhere that's a little further a long than my last company, that's reporting some healthy metrics, green profitable metrics, since my team was let go of, so I'm a little sensitive to that… Also honorable mentions would be the cute logo and to be able to call myself a datadog ha
What is serverless computing?
Why use Datadog vs built-in Azure monitoring?
Why are you interested in this specific position?
Power vs simplicity tradeoffs
How do you work with internal teams?