AWS Hybrid Cloud DevOps

14:52 · Watch on YouTube ↗

Transcript 2,551 words · about 17 min to read

Auto-generated captions from YouTube, not hand-corrected, so names and technical terms may be imperfect. The video is authoritative.

[Music] hey how's it going it's Keith Townsend from the CTO of visor comm with another CTO dose we're actually in Las Vegas I know you can't tell by this angle but maybe you can tell by the weird wallpaper that you get in OKC Las Vegas hotel suites we're at the Cosmopolitan and I'm joined with Chris Williams of green pages technologies you're a cloud architect there yes I am so first off what is green peas green pages is a we traditionally we were a var a value-added reseller but now we've also gotten into the professional services consulting realm and we're also a CSP so so a lot of what I do is the architecture and the professional services deploy side of it and then that dovetails in with the CSP so so I build it and then our CSP takes care of it so cloud we're at a w3 event your CSP help help me understand the relationship of AWS because wasn't AWS both like put all of the hit up in CSP types and MSPs out of business you know we're gonna Alex worse to the cloud and AWS is gonna just take care of everything for us wasn't that the the dream a few years back ah I don't know I can I can tell you that things are going very well for us know in in regards to that every environment still needs care feeding and maintenance all right AWS is not in the in the role of doing non-automated care feeding and maintenance there they're good at automation pieces they're they're really good at provisioning and and deploying and standing up things very very quickly in in term in terms of the day to operations and the long and the long term stuff that are very best spoke to each individual customer that's that is that is a area where we shine so Chris you've participated in the community for she's helped out with the brown bag you've you blog what are you seeing like the intersection of because you're in this weird well what you have kind of one one foot in the vmware vsphere community you know Vemma rah rah rah and then your full head on in AWS community is well what are you seeing that kind of it as the two communities and disciplines murders yeah that's that's actually a really good question because I've on one side I got V brown bag envy tog hi al and on the other hand I also helped to run the AWS Portsmouth user group hey the other guys and and so the the two different user groups that were that we're servicing and that we're working with are they're they're very different right and yet and yet we've got to get them to come together in that magical unicorn of DevOps kind of a kind of role because with with V tug and the beat round back stuff more and more the V tug it's it's the the VMware guys it's the provisioning and infrastructure people and and the operational day-to-day stuff and and in the in the AWS user groups and again also with the V brown-bag because that's kind of a hybrid it's a lot of it developers mmm a lot of a lot of DevOps coming from the developer side of things and and so there's there's there's different needs but both sides of it kind of understand the need to come together and figure out what the other person is doing with with that you know that to pizza team terminology and and you know agile devops and sec ops kind of stuff so from a crossover perspective let even crossover perspective I guess there's this you know Venn diagram where you where you know you have the the code jockeys on one side the guy if you just want to you know my phone is a great example that he went to vmworld a couple years ago did the first hackathon and he was like what kind of hackathon is there that the provision firewalls I'm like I'm not used to this concept exactly so you know you know that was kind of the the cold Jackie's heading head-on into the Ox guys yep yep where do you see the community kind of coalesce into now as we've I think we're a couple of years into this VMware embracing AWS and the VMware community as an extension embracing AWS why are you seeing kind of the the settling points where the infrastructure traditional infrastructure guys understand their role versus cold jockeys and DevOps guys understanding their roles yeah what has that not happy yet it's it's happening and and this is gonna sound very trite but it's all about communication mm-hmm it's it's the the they they are still in their own camps per se but they are understanding the need to instead of throwing the issue over the fence to to to work together to get and and and also to gain some skill sets and in the other side so with V brown-bag in the near future we're going to set up a series on how to how to learn Python how do out and and so we're talking to developers right now about how to get a a learning track going but gear it towards operational people people that already have a good IT skill set but that but that need that extra boost of okay now this is how you read code this is how you understand and smell code and and figure it out and and conversely we're gonna then you know in the future look towards how do we get an infrastructure track but gear at towards developers how do we how do we teach the opposite side of that and and and that that is the the magic sauce that I don't think anybody has really sussed out so that that part you know when you were talking about infrastructure folks learning cold and that kind of I was with you I'm thinking yeah I've been meaning to pick up a Python book the past two years I can see the value in learning cold yes but their lights go the other way around and that's a you know developers learning infrastructure which was it's absolutely good I think through kind of the it seems like there's a higher barrier of learning infrastructure than it is for infrastructure guys to learn cold because we don't have to learn that much cold infrastructure for I don't know let's let's give it a good example that he gave me a good example last night when I ran into you downstairs at the in the lounge you you wrote this script to do what so so we have a customer and what they're doing is they're their provisioning their software solution as a SAS offering hmm so I wrote a CFG a cloud formation template that basically stands up an entire environment with the devstack a UAT stack and a load balanced multi AZ production stack with back-end databases front-end load balanced environment and VPN connectors and V PC to V PC pairing connectors for four attachments into customer environments and and administrative environments so so I built that I wrote that out and and now they're operational people and their development people are taking that script that script that CFT laying down a foundation per customer and then both the ops people and the deaf people go in and manipulate their pieces of it and then eventually the the goal is to then have them modify at the CFT and and have the CFT automate all of those processes so the ops people need to know the code to make the automation and the developers need to know ok if I make this change how does that affect the environment and therefore my ops perspective how much time are you saving ops people so beyond the obvious a curated repeatable process that every time I need to do this there's a consistent platform that that reduces the cost of support etc but from a practical perspective how much time are you saving that customer in that ops team who wouldn't need to go in and recreate and manually do these processes and steps over and over again well so for this particular customer the the initial environment provisioning piece and and that's you know from from the from the time that they're signed off and the time to the time that it's stood up it went from 3 months down to 27 minutes Wow yeah yeah but but the caveat there is that there was a lot of on-prem stuff that needed to be done right on the customer side so so we've taken all of that we've put into the VPC the virtual private cloud and and now that's now there's a 27 minute process from from the time that the CFT gets kicked off to the time that you can actually log in and manipulate the environment so that's a value of ops folks learning how to code they're not you know doing anything differently than they did for me infrastructure perspective before other than software defining what they need to do yep you know now from a development perspective give me an example of where developers benefit from using from learning infrastructure and applying their development steel to the other way around that's the part that like conceptually I get cross-training is a great idea it's good for it's always just good for developers to know infrastructures jobs so that they make infrastructures job easier and they build better applications but gimmies like put some put some meat on the bone for me like what's it how is that practically a good thing I'm glad you asked that Keith put my develop your hat on so so the this is this is very stereotypical this is very broad strokes but but I've seen it happen time and time again so it's so it's legit the when when a developer writes code for a thing they very rarely take into consideration okay this is how this is how this is going to affect the infrastructure in in a normal on premises environment that's not a big deal right most of the time unless unless you're just you go from like you know if you've got like a tiny pipe and you're trying to just jam a whole bunch of data through it then then then it starts to hurt and you try I mean you know you know bus contention and stuff like that that we that's that happens a lot but you don't notice that when it's on premises notice I said premises the in an AWS environment that is hideously expensive you can you can there's one example on from from the from 2017 where a customer who miss wrote a lambda function had a twenty five million dollar per month bill by going in and refactoring it walking with their infrastructure guys fixing fixing the amount of data being sent per packet via that via the application they went from 25 million dungeons like 13 million or something like that it's it was it was a huge cost savings and it was based upon the developer understanding how infrastructure worked yeah I think this is one of those things that I've heard time and again in my experience is that once you give what seems like infinite resources to developers they will use those infinite resources because they have no regard for the backend right you know other problems I've seen is where you know expecting you like Mantha they would create lambda functions that is scale to you know millions of operations per minute or thousands per second whatever the rate is and they overwhelmed the back-end database right yep so they you know it's one of those things that killed the back-end database front-end is that back-end everything goes out crazy so it's one of those things you're really not even from me technologists from a technology perspective but from a business process perspective you need to understand where there's value and maybe it's value that's going up that back in or maybe a smarter to shrink the front-end and cap it in if if it doesn't generate more revenue or more opportunity for revenue why why increase the performance unnecessarily we just thought a tell us we're normally faced with on premises that's we're not usually you know no one normally tells us we're going too fast yeah and and that's one of the things that we're seeing a lot of now is we we're getting into the business use case requirements conversations far far earlier now and and capturing the the requirements what we'd like to do is we'd like to have like a one day on site where we have everybody in the room we have the business guys in the room we have the technology guys in the room we have we have the CTO in the room if possible and we when we advise their no and and so and we hash it all out right with everybody in the room okay what are your requirements what does this need to do and how expensive can it be and and is or is that even a factor and bye-bye by having the cost perspective not just be okay server X cost this much but also the pipes in and the pipes out in the conversations that that thing is having and who it's having with and you know how much money is it making versus how much money is it burning the developer has Tan has to be very very cognizant of what they're going to be doing and how they're going to be writing it and making sure that they don't blow it up and just you know consume all the resources all right so if people wanted to know more about green pages yep how do we fight yet we have a Twitter handle green pages i.t we have the website green pages calm hit me up on Twitter I'm at mr.

wire calm or they're just yeah mr. M is Twi Ari and I blog at mr. wire calm and you know the there's there's there's so many so many things it's many social media aspects you know the AWS portsmouth user group if you're if you're in the New England area we also have the meetups and then we also have V tug yeah yes a lot in this honestly brief v Brown Bag which we've talked about on a program before so you can visit the brown bag and learn how to learn how to code Python C we just come level you can have some content out there on V brown bag until the next CTO of dos will catch you on the Twitter's you can catch me at CTO advisor and the CTO visor comm was where you can find the videos and other great content talk to you next CTO DOS