Life as a Developer on Oracle Cloud!
HackerRank
·
Intermediate
·4y ago
Key Takeaways
The video discusses Oracle Cloud Infrastructure, its impact on the cloud revolution, and the benefits of cloud computing, including agility, scalability, flexibility, and cost savings, with a focus on developer adoption and usage, as well as the use of AI and machine learning services in the cloud.
Full Transcript
[Music] [Music] [Music] [Music] [Applause] [Music] hello and welcome to hakarang tv this is your host adal bandukwala it's a lovely evening here in bangalore india and it's so refreshing to see all the developers tuning in even at this late hour for those of you who are in the bay area and pacific time zone a very good morning to those in the eastern time zone good afternoon and good evening to all my developer friends from the subcontinent in asia it's fascinating that we have close to four and a half thousand people who signed up for this particular live stream i can already feel the wife i can feel the energy i can feel the adrenaline where is that coming from it is all coming from all of those comments that are coming in from linkedin from youtube from twitter from facebook wherever you've tuned in from welcome and like tarun says let's start absolutely tarun will bang on the money and we're starting right now uh akshay good good to see you the shank good to see you shanky good to see you uh i can't quite pronounce your name because i don't understand that language but thank you for joining us wherever you have from it's good to see you as well um it's so nice to see all of you i said i can literally feel the energy the adrenaline and the while so tell us what part of the world are you tuning in from drop in a comment so that we can get a sense of which location country city you're tuning in from and i'll try and feature some of your comments on screen so that we can show dan what kind of an audience we've curated for dan particularly we're delighted to kind of introduce to to you dan garrity svp of developer services for oracle cloud infrastructure dan joined oracle in jan 2020 to build and lead the team whose mission is to drive developer adoption and usage and enable building deploying and operating workloads at scale in oracle's cloud this mission includes cncf influence services like kubernetes streaming lagging events and apis before oci dan spent five years at aws building platforms a developer most of his life he founded and led several startups that ultimately took advantage of cloud services offered by multiple clouds he has an mba and ms in ee computer systems and a bs in ee all from stanford university and has over 60 patents i kid you not that's 6-0 he has 60 patents in tech ranging from coin discrimination through cpu instruction set design dan thank you so much very grateful for having you thank you for joining us on hakan tv i'm super happy to be here love hacker rank i've used it for years and just super happy to be here you made our day by saying just that dan thank you we can't wait to kind of get started now with that we also have with us today hacker ranks co-founder and cto harishantan karunanidhi hari is right knee with ya both of us are in the bangalore office today every time we have a live stream it gives us the opportunity to go back from office it also ensures that our broadband works properly and we don't have the glitch of power going off or potentially broadband getting cut when we are at home so hari thank you so much for making time like always pleasure to have you thanks asus super super excited for the conversation today all right so let's see where all people are tuning in from so i'm going to feature a few people right so we've got someone from iran nafis welcome thank you for joining us good to see you all the way from iran uh let's see we've got abdul who's from bangalore we've got caroline who's tuned in all the way from wales good to see you caroline we've got uh chaitan from pune we've got trixie uh thanks for joining in trixie good to see you we have our good friend jesse jesse hello good to see you good to see you tuning in to this livestream as well jesse is tuned in from vancouver canada we've got uh anthony who's coming from kenya we've got ali ali good to see you from turkey my friend i'd love to get to turkey someday i hope to heard the kebabs that are delicious we also have someone from nigeria muhammad good to see you from nigeria muhammad you've got sagar from bhopal we've got ismail from morocco we've got ahmed from egypt we've got adria from barcelona we've got vincenzo from england man this is pretty crazy girish coming in from us this is pretty much insane you've got someone from greece we've got roger from spain we've got david from peru it's fascinating thank you so much everyone for tuning in please continue tuning in we'd love to kind of continue the conversation with you all right let me just give me a second and we get started with this particular live stream okay uh we're just trying to restore one particular connection all right we're good and i'm gonna jump back so the theme for today is life as a developer on oracle cloud dan love to get started with the first couple of questions for you uh we've curated these questions from our community and we've also tried to kind of prep and see how best we could make use of your time so enterprise cloud spending is projected to increase at a 16 percent cagr cumulative average growth rate between 2016 and 2026 which is this particular decade we are in which indicates that businesses are no longer looking at the cloud slowly as a tool but more like a platform we'd love to know from you then as a cloud veteran how is the way businesses leveraging the cloud changing that's a great question and you know when i hear that question i can relate that to the the gardner study uh also that says that 85 percent of all organizations will embrace a cloud first principle by 2025. that those numbers to me are shockingly low believe it or not i mean i guess that's the problem is i've been in the cloud too long you know but i i honestly believe that you know as corporations and large and small begin to see the agility and the scaling that they can get using the cloud any any cloud uh they're just gonna flock to it it just offers so many advantages over trying to build your own data center and all the you know there's a lot of transformations that have to happen and i can talk about that in a bit and so it kind of makes it a little bit harder for some of the some of the larger organizations to do this but the the benefits of the cloud are just too compelling they're they're gonna take over i mean the agility the scalability the flexibility the the cost savings these things are all gonna drive everyone to the cloud with without exception so i think those numbers are low well that is quite a revelation just when you think like those numbers are insane here comes an expert who says that those numbers are low which pretty much shows why this is the time to be a developer and develop the skills that are required to foster business in the cloud you know just following up on that comment i think most people um most most people think that people go to the cloud for the for the cost savings um and i think maybe some do but once they start developing anything in the cloud and they realize how quickly they can make changes or deploy a new service or oh you need a new database you press a button and you've got a database oh you want to try using a stream or a queue or something like this you can just push a button and have it happen and so what happens is especially the larger organizations that think that they can somehow harvest a cost savings to go to the cloud so it's usually like the financial the chief financial officer that says hey we should move to the cloud and then you get there and then the developers are going wow this makes my job so much easier and we can deploy new releases of our product every week or we can you know we just have so much more agility so i think the drive is economic but the benefit is agility i think that's very very well put dan you know let's get a little bit into the business right the cloud has transformed multiple businesses by improving information efficiency helping them reduce costs streamline the entire software across enterprise mid-market smb large organizations all possible kind of businesses with your experience is there any distinction in the way large-scale organizations enterprises and smbs use and spend for the cloud wow yeah that's that's a really good question actually and i think to a certain extent uh leads a little bit to why you know oracle is a little different than the other clouds and and i'll explain that um so i think yes we do see differences between these sort of classes and we kind of have maybe three classes there's maybe the large enterprise customer which of course oracle is very familiar with because they've been using oracle database technology for 20 years so we we know these people pretty well financial services all these large organizations that have serious transaction processing needs so there is that sort of large enterprise customer then what you're talking about here smb the small and medium-sized businesses they typically are taking us a sas approach as they as they grow their business and so they're a little bit more comfortable with having stuff in the cloud and then you have small companies that are typically born in the cloud as they would as we would say and so the major differences that that we see there it really has a lot to do with um security and compliance and so you and also uh those transformations that i was telling you about so first let me talk a bit about the security and compliance aspects so these larger enterprises you know they've got databases with people's pii in them with people's credit cards in them and they in order to do this they have to be they have to meet all kinds of standards and they have to be certified by all these different certification companies uh to be compliant with specific laws and regulations so in order for them for example to move a database with pii into the cloud they have to make sure that they can retain or gain a new certification or retain the certification that they already have so that's a little bit of extra work and also they want to make sure that they pick you know a cloud provider that has uh the capability of that can make that easy and so we see a lot of larger enterprises hanging on to their existing data centers keeping pii and credit card data safe within their own perimeter and then for them to move to the cloud it is a security and compliance effort small and medium-sized businesses already have their data in the cloud they're already using sas suppliers for example to process their credit cards they're already in the cloud and so for them it's much easier to extend however they're typically using sas services already so what they want to do is extend their sas solutions that they have so that's a little bit of a different uh approach to the cloud born in the cloud is of course born in the cloud and what we find with them is they're often not worrying about security and compliance until they have to worry about it and so that's one of the reasons that you know we spend a lot of time and energy making security compliance easy so when our developers come in and develop and i don't mean this to be an ad but when our developers come on and and use oracle cloud we basically make them look like heroes in front of the security and compliance people and i can talk about uh why in a minute but let me backpedal a little bit and just talk about the three transformations that have to happen to move from software in a data center to software in the cloud when you have software in a data center you're buying hardware for example on a two or three year cycle you've got a capital expenditure budget you go through all this planning you order the stuff it comes you've got people in the data center that configure it and all that financially speaking all your data center expenses are capital expenditures and when you move to the cloud that all changes instead of having one or two people making all the decisions about what equipment you're going to buy you've got all your developers out there making decisions every single day about what to spin up and what not to spin up and this drives many companies uh insane right because they're so used to having one or two people they can just call them and say hey what's our cap what's our i.t spend next year and they already know because they've already bought all the hardware in the cloud that all changes and it becomes operating expense so that's one big transformation the second one is of course the technologies that are available so if you're a corporation and you have your own data center and you're running windows everywhere you know you're going to have a certain set of technologies that you're going to be developing for when you get to the cloud all of a sudden you can use anything you want any brand of unix you can use windows of course you can use macs you can use anything you want because it's all out there so that the technologies change and and you're also using things to communicate between microservices or you're using kubernetes and you're using service mesh to wire it all together and you're using cues and streaming to link your application components together it's a lot different than when you're developing for a data center um and so that that's the second one is the technology and then the third one is you have to change the way you develop software and this is moving from you know waterfall where it used to be you know design code test and you do that over a period of three months or six months and you do a release every three months or a release every six months now in the cloud you're doing ci cd you write some code you commit it uh it goes up for a code review to another coder they pass it it goes to production i mean it goes through all the stages of course it has to go through testing integration testing etc but you know you're you're pushing code every day out to production and that's just a very different way of doing software so those are the three transformations that have to be made those transformations are harder for enterprises to make due to security and compliance due to budgets and so forth and so we tend to see small and medium sized businesses uh moving into the cloud using open source technologies they're they're less afraid of open source technologies there's less security and compliance issues around open source so they're out there the enterprises it's very deliberate and they move applications whole lockstep awesome thank you thank you dan for giving us the trifecta of what's uh catapulting the cloud onto a totally different level these days now i've got a ton of questions i want to ask but then i know hari wants to ask you a few questions as well and a lot of those questions might also want to come from the audience so i was practically counting the number of countries that people have participated in from today manually i saw that number is north of 30 which is incredible so you've got participation from 30 different countries today so if you have any question that you'd like to ask dan or hari for that matter please feel free to comment that question and what i will do is then i can pick up that particular question and flash it on screen just like this hey kiran good to see you hello so just like that i will flash it on the screen and then we can go ahead and ask um ask dan and dan will be happy to answer so uh just like for example chirag is asking what is cicd uh before dan can answer it's continuous integration continuous development then if you can confirm uh that is correct or incorrect bonus points i'm sorry i i i always tell people at my own company that are speaking don't use acronyms because you know not everyone will know but cicd is yeah this process of continuous integration continuous development what that means so in the old days remember you'd write some software and you'd commit it to rvs or whatever version control system you have and then every night you do a build that's the integration is the build and then if you failed if the build failed because of your code you had to come in and i don't know we used to put a dollar bill in a jar and then you know at the end of the project when it shipped we all enjoyed a nice party based on on who could build or not well in continuous integration what happens is every time you make a commit and depending on how you set it up but i'll just do it how we do it anytime you make a commit another code reviewer gets a message and a different engineer looks at your commit and says yes or no and if they say yes that commit is merged into the master branch and it's built right then and there the software is built right then and there and run against all the integration all the unit if this is the unit test it goes on to the integration tests if it passes all those it goes out to an alpha environment where it's live if it passes all those tests it goes to beta and then it goes to production so that's the continuous deployment part the continuous integration means when you commit your code it builds right then and there so every time somebody does a commit your software is rebuilding itself all the time continuously and then it gets pushed out all the way out to production now in order to do this you have to have super high confidence in your unit tests and integration tests because you know there's no gates i mean the gates are are automated so like if your software passes the test it's going out to production so you kind of want to make sure you have good tests in place and that's why the whole practice of cicd first of all it well it improves velocity because seriously you make a change right now it can be out in production in the wild in six hours like you don't have to wait for any devops team or any deployment schedule anything it just goes so again getting back to agility but the other thing it does is it really does encourage uh developers to um do test driven design i mean it really you really do have to have some good testing frameworks and some good ability to do to do load testing and distributed testing and all these kinds of things because you got to have the capability to test to put in ci cd so that's ci cd people you know some people don't make a distinction between ci and cd some people use slightly different words but that's why we just call it ci cd it's kind of the whole pipeline from commit to production brilliantly said uh thank you so much dan and chirag so there we hope you got it it was very well explained by dan what cicd is and then if anyone else has more questions feel free to comment just like that and i'll feature your question on screen and we can ask that that in the meanwhile i'm picking up common threads so i see that there are at least 12 to 13 people who've asked that same question how can you potentially join oracle i think that's a rather simple one to answer orton is soon going to be conducting a hackathon and it's coming up on screen right now so if you're keen to be part of oracle's cloud if you'd like to be notified about their hackathon which is coming in soon you can go to that particular url that is on screen once you go to that particular url you can sign up you will be notified as and when the hackathon is live participate in the hackathon and then there's your chance to be part of oracle cloud uh with that if you have any other questions do type in the comments otherwise i'm giving handing it over to hari i i know hari you've been itching to ask those questions thanks adele i've been waiting for uh uh you know to start with some of the technical questions i've always wanted to ask and um then first i want to start with uh you know how you're using ai and and ai as a technology has revolutionized multiple sectors without any exception and cloud computing is no stranger to it as end users we have just you know launched say containers or use cloud to build ai i would love to know how you know more insights on how ai is leveraged by oracle cloud like how does cloud use ai to build cloud like can you give more insights there well sure uh yeah we are using so so artificial intelligence or machine learning and again just using uh just explaining acronyms and most people just say aiml um and usually people kind of distinguish those two things but for this question i'm just we're just gonna call it aiml and oracle has had a long history of doing aiml mostly again it's the the database legacy so they put aiml services right into the database in in well they're still there today but they've been doing that for a long time so now we've taken that and and put it into the cloud and and every cloud of course has aiml capabilities and they make those uh services available to developers just like you said so if if someone wants to do for example sentiments analysis they want to understand they have comments that they receive back and they have 30 000 comments and they want to sift through and see how these comments positive or negative they can just call it's it's literally a service called sentiment analysis they can just send that data in and they get a json back you know good bad ugly um in terms of sentiment analysis um but internally we are using the same technology to to help our developers so example um ci cd so if you're doing ci cd we just talked about how you need to kick off a build every approved commit and so how do we do that build i mean do we use big virtual machines or little virtual machines or do we use uh you know how much memory do these things need and all of these things and so what we do is we watch your builds and we adjust the configuration of the machines that we're using to do your build according to your preference so you put a little dial in you say i'd like my builds to be fast okay that's going to cost a little more money but we'll optimize for fast we'll do more parallel parallelism boy more parallelism and potentially we'll use bigger vms it depends on you know if you're building java we know how much memory it takes to do a good java build etc so we use ai to first yes what you're going to need because oh you're using java and you're using these dependencies and you're using this cache and so forth um and then at we watch what happens and then we tune it and so basically every build you do gets better and better and better same thing with kubernetes so we're really starting to observe these patterns that people are doing with kubernetes and you'd be surprised at how many people are kind of doing a similar thing right they're looking at either cpu or memory and they're wanting to expand or you know to scale out their infrastructure or scale it in and this is something that kubernetes which is a container orchestration system something that kubernetes does really but you kind of have to tell it what you want and so this way applying ai now we can move to a managed scale out and that's very helpful for for developers another area that we're using and this this one is really actually pretty cool i mean there is nothing worse than being in charge of patching and the reason i say that is because it never works like you have a gazillion different versions and flavors of operating systems out there you get a new release you need to apply this operating system release and often you need to reboot the machine and when you reboot the machine the services come up in a slightly different order and something doesn't work and so it's really a pain in the rear well oracle's been doing a lot of work in this area and we have for example oracle linux which can do an in memory patch um but also what we're doing is we're looking at for not just the operating system but also the java runtime and some of the other runtimes but i think java runtime in particular we know what you're using we know what the security threats are we can actually figure out which patches to apply to your stuff in which order so that we don't disrupt your service that is a really cool use of ai on the cloud side the developer never sees this all they know is that they don't have to patch their machines or they don't have to reboot as they're patching they get the advisory they can decide whether they want it automatically applied or or on request but the answer is the same so we're using that ai technology to simplify the life of the developer either automate patching automate scaling automate building all these things but doing it against a hard problem so you need a whole bunch of data to figure it out all right thanks for that and i would have predicted scaling and building to be automated but patching was never in the list of things i ever imagined ai is going to go and help i thought it's a standard boring thing you take it apply it hopefully it works pray for it and then you know you are going to go to the next version the next version i think it's the prey for it that is the killer you know because we have to patch obviously we patch you know all continuously and um it's uh it's it's everyone's challenge it's a challenge in the cloud yeah it's it's amazing to see ai take over that and and talking about the technology of future uh there is this with web three and and you know right now uh decentralization gaining a lot of prevalence do you think like cow club cloud computing faces thread a thread from something more decent centralized like distributed computing like is this is is the web3 gonna help growth of crow cloud or is it is it against the cloud at this point well so i mean i think it's judo which i'll explain in a minute but first i just want to make sure that our listeners understand web 3. i'm just thinking that if if logically if people are saying you know cicd maybe we're trying to learn something about web3 too but the idea behind web3 right is to push the computing out to the data to push the computing out to the edge to use zero trust between machines so meaning the old pki certificates aren't necessarily going to work and you're trying to distribute the ownership and distribute the monetization of the web now i no i don't see it competing um so the way that judo works is you know you use the force and the energy of your opponent in your direction so if somebody comes and lunges at you you you take advantage of that so i think that's really what's going to happen with web 3. one of the things that that oracle is doing and and again i i mean everyone's doing it so i'm not this isn't an oracle ad but i think we're doing it a little bit more aggressively and uh and certainly more completely than than our competitors but that's moving the cloud out to the edge moving it all the way into our customers data centers so customers again are reluctant to put their pii and and other sensitive data in the cloud they want to keep it in contained in their data center now oracle has a 12 well i i shouldn't say how many racks it is but it's a small set of racks that you can put in your own data center and you have the oracle cloud in your data center the whole oracle cloud not just compute and storage you know and a few other things like this um but it's actually what we call a dedicated realm cloud at customer we call it drcc so that's the ultimate expression of computing on the edge because you have compute data database storage networking on all the ci cds or everything you have and you can take your cloud that you have in your own data center and you can wire it up to the internet in which case it's just like oracle's public cloud or you can isolate it from the internet in which case it's your own private cloud and we've proven this technology out now in the us government and multiple occasions where we have a complete cloud and we're able to still update that cloud we keep the software everything running with the latest releases we get our four nines of availability and and yet uh these people get to have all these services with with microsecond latency and inside their own uh firewall if you will um so i think what we're seeing is a is it's i think web 3 is a is a trend that the whole cloud is going to move to i guess that's the way to answer the question and i don't i do worry a little bit about there are some lofty ideals in web 3 like you know the financial ownership and like blockchain basically everybody being able to do everything anonymously and i think that's good but i think ultimately the reason that there are only sort of four u.s cloud providers and two chinese cloud providers is that it takes a lot of capital to do it right and it's gonna and and those people that have to spend the capital are typically the ones that get the remuneration so i just i don't i worry about that part a little bit but as far as being distributed zero trust anonymity no certificates yeah i see that happening that's good and and and having the entire cloud in the private data center i think that's the right example of best of both worlds like you have everything uh and and you know you have cloud in your setup and that that's leads to the next question i wanted to ask which is oracle is a little new to the cloud game uh with its platform established in 2016 compared to you know aws in 2006 and google cloud in 2008 uh you already mentioned one love to know more on how is oracle differentiating itself from other established players in the space and what what is the focus on with respect to hiring for oracle club sure um well let me take your first question first about like established in 2016 and it's true or oracle's cloud is established in 2016 but we have worked so quickly if you line up 2016 with you know 20 2006 for aws and then you look at the growth and in revenue and services and all the things that they have we are completely outstripping them it's not to say it's going to go on forever but i think that because we're a second fast follow if you will it's not even that fast because we're a second comer we've been able to avoid a lot of the challenges that i think the the other clouds have had to deal with and and and the first example of that is i just explained about how we can have a drcc a dedicated cloud region in someone's data center well we have to build that you know you install the hardware but we have to create the cloud within that hardware this is something that is really hard to do i it's it's almost impossible to describe but you can imagine um that for example storage meaning like let's say attached storage so a disk drive it needs compute to be able to to build its service well compute needs block storage to be able to build its store its sync and so they depend on each other and so you can't you can't just sort of say build a then b then it's more like trying to do a package management in in unix and having to be hopelessly convoluted and so from the outset we knew we were going to have to build these regions um with a button press and we're not quite there but we're darn close so when we want to build a region we push a button it creates a a virtual uh environment and we build out the infrastructure the core infrastructure then we copy that into the new region then all the software services come on in a specific order you know they advertise what capabilities they have advertisement capabilities they need and we're kind of applying a little bit of our own yeah ai to try to figure this out and we built so we're the only cloud as far as i know that can do that they can push a button and build a region and because of that we have a lot of regions we already have 40 regions i think i'm not sure what i'm supposed to say but we have a lot more coming and oracle's strategy so uh i think aws for example likes to build large data centers that have three availability zones and those are uh separated by 100 miles and this kind of thing well oracle's customers are more hey they want to build an app in one region and then they want to have a dr region nearby in the same typically the same country but the same zone at least and so that's oracle strategy really is a lot more regions and regions that really have a primary and secondary site in the same country and so that's how you'll see oracle spreading out and again this is for people for customers that want to have um pii european customers wanting to keep their their customers private information in their country you know this is we call these sovereign clouds and there's a lot of of uh countries out there and a lot of businesses out there that want to keep their data within certain geographical boundaries and we can do that better than any other cloud so we can put it in your data center if you want or we can put it in a nearby city with a successful dr plan um so that that's one thing that we're what what we're doing and then i think the other thing is that oracle is big on hybrid like i don't think we have a single customer that's not well i shouldn't say that that's a terrible thing to say i think many of our larger organizations are running some on their in their data center and some in oracle's cloud and we even have a connection with uh microsoft's cloud a super high speed interconnect so if you want to run a net application for example against an oracle database you can do that you just run i mean so not only are we hybrid between private data center and public cloud but we're also hybrid cloud to cloud so i think that that's something that's unique for us again i'll go back to the security and compliance we have uh well it's we have a centralized control plane and i don't know how deep people want to get into this but every internal service you basically uses an internal uh api gateway and so we're able to provide all the governance features an api gateway you know throttling uh you know tagging we do tagging we do logging we do so basically no matter what the service is we know who created what resource when and so this and this data is available to the customer free in this you don't have to do anything and so it's makes it really easy for a developer to build something and then again just go to the compliance officer and say here's here's your log here's what everyone's doing and it's the whole cloud nobody else has that so i mean we also offer some migration solutions i mean particularly for database users so i think you know other clouds are like thou shalt not use relational databases we're obviously not like that because actually relational databases are a pretty damn good tool um and so we support that and we support some migration from on-prem to to the cloud and of course we have oracle database which is arguably the world's highest performance database but we also have my sequel arguably the world's most popular open source database so we've got all these different things uh to choose from so and then finally price performance we do have like our high performance computing for example we still have bare metal available so that's pretty unusual you can get some pretty awesome performance uh out of the oracle cloud and it is it's affordably priced so with respect to hiring um you know i think that the pandemic has changed that a little bit um we i have my team now i have uh i don't know a certain number of people but i think i'm in like 55 cities in 22 countries right now and we were kind of trying to centralize these things and trying to kind of create centers like in bangaluru or you know we have a our india development center down there we've got a lot of of people there and we're hiring pretty aggressively in india but and then when the pandemic hit now it's much more we're pretty sure that people can be uh can work from anywhere and so uh it's much easier we can hire from pretty much anywhere in the world um so i think i'm not sure what you were curious about from you mean in terms of like hiring how do we differentiate ourselves against our competitors from a hiring perspective is that what you were asking yeah most of some of how some of the differentiating things you're building uh how does the how does that affect the you know the technologies you're hiring for or the skills you're hiring for and levels you're hiring for like how does the new things you're building for example i'm still when you s the drcc or the clouded customer it's one of those things when before you build or when someone pitches that idea the right reaction is like are you crazy and then once you build it it feels like magic like there is there is no other way to describe it like i'm still i'm still amazed by the amount of complexity which can go through in that and to hire for something like that like how does what it's funny because because you and i have had this conversation where i've sort of said hey you know i wish that hacker rank and we can talk about this again offline but um you know there is a lot i mean the fact of the matter is and whether you agree with this or not or like it or not the fact is is that people have to be able to write automation software so like it's not just about writing the java on your laptop and watching the robot go around like i mean people do need to write hardcore java multi-threaded distributed systems uh but then they also need to write code to deploy that system to test that system to make sure that system is able to be up and healthy after a after a patch to make sure that that system is operating at four nines of availability and a lot of times people come in thinking okay i've been hired as a software developer at fill in the blank cloud this isn't true just for oracle i mean you know somebody said oh i've been hired to to develop you know java or i get to write rust for uh aws and i'm gonna design all these cool applications and you do but you also end up writing automation software in python or in bash script or in some other tool and i think that people are not really exposed to that so certainly if somebody has some sre experience or has some automation experience they're going to be a much more attractive hire for us good got it my last question is is uh you know when cloud has its own challenges right like the cloud has changed the ways in which we build deploy software but comes with this challenge and one is when the cloud has its own outages right like we have seen and when the cloud goes down it takes multiple services along with it and can you you know give us more insight on if there are steps being taken to mitigate these problems either from the cloud or the platform side or from the customers on the company said like what do what do we need to do to you know reduce the risk from the outages yeah you know so good question um i think with respect to our cloud for example there's all kinds of things we're doing to get to four nines you know we want to be four nines it's you know four minutes of outage in a month something like that um you know and sometimes even that's not acceptable to a customer i mean certain services for example dns have to be five nines or six you know it just we can't really tolerate that kind of but availability is on everyone's mind and the problem is as as we've been talking about this inner interdependency and interconnectivity and this depends on that which depends on this and so often people don't realize that their software is vulnerable to for example a dns failure or for example a block storage failure or a networking failure so i think um you know what we're doing about that is understanding our dependencies and that's part of the work that we have to do to build the region anyway so we're just extending that into operating the region so if i know what i'm relying on my service if i honestly know what i'm relying on then i can build in contingencies so you know what we can i do think that you know oracle's cloud is high availability high reliability um along with the other leading cloud providers i think you know for customers it's uh it's a challenge so i think if you're an aws customer you know often you get the advice hey there's three availability zones in a region take advantage of that make sure you're deploying in all three of those azs because then if one goes down you still have two um i i think oracle customers are more likely to do a primary secondary then a primary secondary with a failover and i'm just saying this is what we're seeing i'm not saying that one is better than the other but i think you know as a customer you kind of have to understand those two different approaches if you do rely on the 3az approach yeah if one az goes down you're probably going to be safe but as we've seen sometimes the whole region goes down and personally i think that the customers that sort of set it up with a primary secondary failover are in much better shape because uh the whole region maybe only part of the region goes down but they still can fail over to a secondary site now maybe that's more less cost effective i don't know you can't you know and we also offer fault domains within our regions so but these are not separate data centers like they are at other cloud providers these are separate power supplies separate racks separate you know if the if the data center goes offline then you'll still lose it so we you know i think most of our enterprise customers are going primary secondary got it got it i mentioned that's the last question but the the developer side of me i do want to ask one more question which is what is what is life like for a developer on oracle crowd cloud what does their career look like what what roles are you know changing the game and water aspiration aspirational in the world of oracle cloud like what if i'm a dev working in oracle cloud what kind of challenges will i be solving there developer working like working for oracle oracle club yes yeah um well we we have a lot of great challenges i mean building these distributed systems is uh really insane i mean if you're a developer and you're in school and you're learning about these the cap theorem and you know and guarantees about exactly once you know you know only at least once exactly once or you know these things are big big big issues trying to do distributed locking trying to build services that span uh regions trying to build cross-realm replication when you have a completely different set of secrets in one realm and another like these are big problems and all of the cloud providers have them i i think um for us you know with so many regions we tend to have a lot of uh challenges with these yeah distributed systems that are distributed across regions so i think really if you come to oracle you can come out of college you can come as more like an ic3 with you know three four or five years of work experience and the first thing you're going to get into is you know highly scalable highly distributed watch how you know what what is your thread pool and are you paying attention to that because you could exhaust your thread pool or if you're running node how many processes are you running because you know you can get exhausted and just having to look at all of those pooled resources and making sure you're alarming on it because these things just grow and grow and grow i mean we we see like our our streaming service this year is up 500 x what it was last year so like if you didn't think about that when you were building the streaming service it's going to be dead right i mean how could 500 x the throughput you know petabytes and it's just crazy and so i think it's really about thinking big about what we call 10x thinking i don't want anyone working on any line of code that can't handle 10 times what it's handling today because otherwise it's not worth it like don't work on something to get a 10 performance improvement so it's really a different mindset it's really about scalability it's about distributed it's about multi-threaded multi-process uh computing that's that hits four nines of availability and that is not easy so it's a great challenge for any developer honestly um so thanks for sharing that i think four nines is about you can go maybe a couple of seconds a day or like not like it it's few minutes a quarter like two to three minutes of water if i remember correctly uh i know either you have a ton of questions from the you know people who are watching this and i know you have a few questions as well over to you sure thank you hari so a lot of questions coming in from the audience there's one from alex mann alex is based in ashland oregon in the united states and this is his question he says dan what kind of growth do you see in security identity management and what are the new challenges these roles will face with implementation of cloud and ci cd okay i think i should take the first part of that question and make sure that i understand it so what are the challenges we're seeing in identity is that kind of i don't know about that i mean that is huge right um that's absolutely enormous and it used to be that kind of the world was like oh well we're going to use active directory and then it was like well we need our active directory to be out in the cloud okay we'll we'll use you know microsoft's cloud active directory that's it everything relies on that well you know it doesn't really work like that and i think there's a lot of identity issues when you get into the cloud first of all people are trying to merge cloud sas and on-prem so you've got effectively three sources of identity right there like you you can't you know and that's just one sas provider you know most these people are using you know such and such and such and such and they in the old days it used to be okay we'll just use sso and you know we'll use uh saml and we'll basically tie to the corporate identity system we can't do that in a distributed system if you have a system that's spanning multiple regions you need to authenticate in those different regions you have different tendencies in those different regions you have different credentials and you need to use you know a vault and you know a secrets vault and then you need the the secret to the other vault in the vault and it just gets very very complicated and what the the challenges are is federating all these things so it really is trying to give one person one place they can sign in and that will give them access to all the different things that they need and this is really a challenging thing to do in identity if if someone could kind of reinvent identity from the ground up they'd probably do it a lot differently so that's what we're doing we have sort of this what we call an identity domain and then you can hang a bunch of different identities onto this one and you can have multiple identity domains in case you're using multiple regions and so forth then you can link them but these are really big problems and they require a lot of thought and a lot of understanding of security crypto this kind of thing so there's a lot of challenges there what was the second part of your question i i missed it you were just thinking about identity yes as a whole yes uh so the second part of the question is what new challenges will those roles face with the implementation of cloud and ci cd oh cloud and ci cd well so ci cd we were basically using vault so vault is a hardware security module an hsm a cloud hsm and so you can store credentials in the vault um safely and so that's what we're doing for ci cd so we're you know because you're deploying to something and that something needs to be either a region or a kubernetes cluster or a grouping of vms or or whatever it is and so that's got to have the credentials we put them in a vault the developer doesn't have to know what the credentials are which is really nice and there's no storage of any credentials on somebody's laptop so so that's how we're managing it but yeah awesome awesome thank you for sharing that done with that i think with just on time i know there were a lot of questions it's fascinating how time has uh gone by it's been close to an hour we've been talking and then believe it or not there are more than 617 comments that i see right now that have come in on linkedin alone i mean you think about twitter youtube and facebook i'm sure there are much more but i was just trying to refresh linkedin and see what kind of comments questions have come in i just saw the 617 so clearly a lot of questions a lot of interest uh very grateful to you uh dan for your time in uh making some time today to interact with our community thank you so much my pleasure hari thanks a lot for making time like always and to all of you who who tuned in from all those 32 plus countries that i counted and i'm sure there are much more thank you very much you have a great day you have a great afternoon and to all of those night owls in india and in the asian subcontinent good night and god bless okay thanks [Music] [Music] [Music] you
Original Description
The Cloud Revolution has been at the forefront of the next wave of software and digital companies, and Oracle Cloud has had a very significant impact in making this happen. ⚡️
In this livestream, we’re going to host Dan Gerrity, the SVP, Developer Services at Oracle Cloud, who’s going to talk to us about Oracle Cloud - their platform, the impact they’re creating, and what’s life like for a developer working for Oracle Cloud! 😀 🚀
Watch on YouTube ↗
(saves to browser)
Sign in to unlock AI tutor explanation · ⚡30
Playlist
Uploads from HackerRank · HackerRank · 0 of 60
← Previous
Next →
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
HackerRank for Professors
HackerRank
Why CareerCup's Gayle Laakmann McDowell Loves HackerRank [Testimonial]
HackerRank
How to Leverage Data-Driven Insights to Improve Recruiting Results
HackerRank
Hire Top Linux Talent and System Admins with SudoRank [Webinar]
HackerRank
How Zenefits Hired 10 Engineers in 10 Days with HackerRank's CodeSprint
HackerRank
Hire the hottest SQL talent with DbRank from HackerRank [Tutorial]
HackerRank
How to Use HackerRank's CodePair Tool [Tutorial]
HackerRank
Find Linux Talent with SudoRank challenges [Tutorial]
HackerRank
VMware's Mat Connot details his favorite new way to engage top programmers
HackerRank
An Inside Look at Coinbase's Recruiting Strategy [webinar]
HackerRank
How Evernote Enhanced Their Tech Recruiting Process
HackerRank
5 strategies for being a tech recruiting champion with HackerRank [Webinar]
HackerRank
HackerRank Story - Lets make the world flat!
HackerRank
Data Structures: Hash Tables
HackerRank
Data Structures: Anagram Problem Solution
HackerRank
Algorithms: Solve 'Lonley Integer' Using Bit Manipulation
HackerRank
Big O Notation
HackerRank
Algorithms: Graph Search, DFS and BFS
HackerRank
Algorithms: Bit Manipulation
HackerRank
Data Structures: Solve 'Find the Running Median' Using Heaps
HackerRank
Data Structures: Heaps
HackerRank
Algorithms: Memoization and Dynamic Programming
HackerRank
Algorithms: Sort An Array with Comparator
HackerRank
Algorithms: Solve 'Coin Change' Using Memoization and DP
HackerRank
Data Structures: Solve 'Contacts' Using Tries
HackerRank
Data Structures: Tries
HackerRank
Algorithms: Bubble Sort
HackerRank
Algorithms: Merge Sort
HackerRank
Algorithms: Quicksort
HackerRank
Data Structures: Binary Search Tree
HackerRank
Data Structures: Trees
HackerRank
Algorithms: Solve 'Shortest Reach' Using BFS
HackerRank
Algorithms: Solve 'Connected Cells' Using DFS
HackerRank
Algorithms: Solve 'Recursive Staircase' Using Recursion
HackerRank
Algorithms: Binary Search
HackerRank
Algorithms: Solve 'Ice Cream Parlor' Using Binary Search
HackerRank
Data Structures: Linked Lists
HackerRank
Data Structures: Cycles in a Linked List
HackerRank
Data Structures: Stacks and Queues
HackerRank
Data Structures: Queue With Two Stacks
HackerRank
Data Structures: Balanced Parentheses in Expression
HackerRank
Algorithms: Recursion
HackerRank
How to Hire for Your Startup [Collision Panel Video]
HackerRank
HackerRank main() San Francisco | November 8, 2018
HackerRank
The Enterprise Playbook for Technical Hiring
HackerRank
Remote Hiring Colloquy with EY's Dipika Kapadia & BigBasket's Rupesh Kumar
HackerRank
Remote Hiring Colloquy with Hackerrank's Hari K, Akamai's Neha Akhouri & Infosys's Rafee Tarafdar
HackerRank
Building a Culture for Developer Innovation
HackerRank
Developer Trends in 2020 and Beyond
HackerRank
HackerRank's Livestream: Accelerating Developer Innovation at PhonePe
HackerRank
Aadil Bandukwala converses with Rahul Chari about HackerRank's upcoming livestream
HackerRank
Accelerating Developer Innovation at PhonePe with Rahul Chari
HackerRank
The Future of Work for Developers
HackerRank
Scaling AI from Proof of Concept to Production
HackerRank
The Secret Sauce to getting hired in Myntra’s Tech Team
HackerRank
Reimagining Money with Paypal
HackerRank
Engineering Predictions for 2021 and Beyond! 🎬
HackerRank
Future-forward Tech & Careers with BNY Mellon
HackerRank
How Atlassian Approaches Diversity and Inclusion with Balance & Belonging
HackerRank
Connecting Global Tech Ecosystems: Andela Shines a Light on Africa’s Developer Talent
HackerRank
More on: AI Systems Design
View skill →
🎓
Tutor Explanation
DeepCamp AI