Cloud Generation 2: Larry Ellison Keynote at Oracle OpenWorld 2018

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
the Oracle autonomous database the self-driving day sir relax everything's automated human error is eliminated a revolutionary new technology called machine learning enables the database to protect and repair itself without ever having to stop the database automatically responds to cyberattacks and prevents data theft in a world striving to build self-driving cars Oracle has built a self-driving database fully automated ever-vigilant your data is safe [Music] in the auricle autonomous database cloud [Music] ladies and gentlemen please welcome Larry Ellison [Applause] hi everybody I'm sure you're curious did I also do the drawings on that last ad the effect is no I was that was a collaboration but okay today I'd like to talk about the second generation of our cloud featuring Star Wars cyber defenses to protect our generation to cloud we've had to react detect it from the ground up and introduce Star Wars defenses impenetrable barriers or force fields an autonomous robots the combination of those two things protect your data and protect our generation to cloud okay the current state of the art of cloud defenses just not good enough not even close the smartest technology companies are routinely penetrated and you can read about how vast quantities of their data is stolen the most security-conscious government agencies are also vulnerable if the smartest among us the most security-conscious among us have been attacked successfully this is a big problem the attacks are getting more frequent but the attackers are getting smarter and our businesses our data is more vulnerable as a result of this oracle has gone through a period of fundamentally re architecting our cloud we built a generation to cloud and we built in again these star wars cyber defenses now it's easy to say very hard to do to build a secure cloud if it was easy to do someone would have already done it we've been working on it for a long time as I said it required a fundamental reaaahhr tech chure of our cloud and we did that and as a result we have these two key technologies that protect the cloud and protect your data impenetrable barriers block threats from getting into the cloud and then autonomous robots find those threats and kill them to want to really check the cloud I'm not talking about a few software changes here and a few software changes there I'm talking about a completely new hardware configuration for the cloud starts with the foundation of the hardware we had to add a new network of dedicated independent computers to basically surround the perimeter of our cloud these are these are computers you don't find in other clouds a whole new network of dedicated computers to form this impenetrable barrier and it not only protects the perimeter of our cloud caps keeps threats from getting in these barriers also surround each individual customer zone in our cloud so threats cannot spread from one customer area to another so one customer a malicious customer can't move laterally in our cloud and see the data and get it the data of other customers that required a completely new hardware configuration as well as new software then we use the latest technology the latest AI machine learning technology to build autonomous robots to go out and search and destroy threats that have yeah that are inside of our cloud now we talked about for a long time we have an autonomous database and the autonomous database is autonomous because it has a bunch of robots doing things it has robots that automatically tune the database it has robots that automatically backup the database they recovered the database when the database goes down shouldn't ever go down but if it you know recovers it if it does the now we've added lots lots more robots to protect every aspect of the cloud and it has to be robots in the old days and the way most people operate today is if a vulnerability is found in your database the vulnerability is found in your applications and your systems human beings then decide how do I schedule a downtime window I need to take that system down and then patch it and by the way I got to find all the related systems if it's a strut system I'm gonna find all the strut systems then patch all the apaches struts systems I got to locate them all see which ones I can take down now which ones I can take down later sketch schedule patching do all of that in other words it's a bunch of people trying to defend your data against a robot or botnet attack it's their robots versus your people who do things faster who do you think is gonna win it's got to be a case where this is completely automated its autonomous a database threat is discovered automatically no human beings are involved the patch is immediately applied while the database is running no downtime window has to be found you can patch this system while it still runs this is essential for protecting our data and protecting our cloud here's the problem the generation 1 cloud has these very simple great price performance Intel commodity computers they're great machines we use them to to socket machines and you can have multiple customers in the same machine no problem well maybe a problem but that's not the biggest problem if you look at the AWS cloud in that machine could be one customer could be multiple customers but in that machine is the AWS cloud control code sharing the computer with customer code that means you better trust trust your customers you better trust all your customers if you're gonna let your customers inject code or use the computer that you use to control the cloud if you're gonna let customers share that computer the computer you use to control your cloud and those customers are smart they can look at your cloud control code they can change your cloud control code they can move from one computer to the other they can look at other customers data they can schedule the other customers data is exfiltrated out of the cloud someplace else and they can make sure that you get the bill twice you pay for the exfil and your data is lost if you have a single shared computer running your cloud and running your customer code one customer can see the other customers data Amazon can see your data and the customers can change the Amazon code and hack the system and take control of the code and steal data it's a big problem it's a fundamental problem with the architecture of the cloud what we have done on the right with our Gen 2 cloud is we have customer computers that the customers can use we call them bare metal computers by the way also I'll show you a little more detailed picture but you can have one customer in the bare-metal computer you can even have multiple customers in the bare metal computer some customers will choose to have that computer all to themselves some customer will say well I don't really care I'm very cost sensitive I'm happy to share the computer with other customers it depends but we will never put our Cloud Control code in the same computer that has customer code that creates an incredible vulnerability to our cloud control system so we've added a completely separate network of dedicated cloud control computers that not only protect the perimeter of the cloud protect threats from coming from the outside and getting into the cloud but we also form a perimeter around each individual customer zone so customers can't get out of their zone and into your zone and they can't hack our cloud control computer because there's no way to access it there is no access to our cloud control computer they can't look at the memory they can't add code they can't do anything to it it's an isolated network they can't get out we can't look at your data which is nice but the most important thing they can't look at our control code they can't change our control code okay so this impenetrable barrier created by the cloud control computers not only surrounds the perimeter but just as importantly surrounds and isolates every individual customers own threats can't editor enter and threats can't spread it's a big deal all right I'd like to clean up some naming there I'll you I'll use three different names and they sound synonymous but they're subtly different I'll talk about our generation to cloud by the way this is a picture of what we call Oracle cloud infrastructure OCI is our second-generation cloud infrastructure look at the top line on the end a box in the upper left-hand corner says bare metal that's a bare metal server this is a server that users can rent that has none of our code on it and the users can have that all to themselves so no other use you're gonna you know if you've got you want you can have the series these bare metal servers only your code on it only your customer no other customers code and certainly none of our code so you don't have to trust us no offense we don't trust you well we trust you but we don't trust all of you if you know what I mean I say we've got a bare metal server and we also have VMs we have Exadata in our cloud infrastructure so if you're running our database you can run it on a shared Exadata as you'll see and I'm going to announce in a few minutes you can also run it undead ik ated Exadata infrastructure where the you're the only one using that Exadata machine you can isolate yourself there so you can isolate yourself on bare metal computers customers can have their own computers their own Exadata machines oh and their entire perimeter sealed off by our cloud control computers other customers can't get look at your data we can't look at your data okay this is this this whole picture is the picture of our Oracle cloud infrastructure our Gen 2 infrastructure what's not shown is the entire gen 2 cloud which includes our fusions things like our fusion SAS apps there's another layer on top of that so when I talked about the gen 2 cloud and also talk about Gen 2 infrastructure slightly different the infrastructure doesn't include the analytic services doesn't look loo to include the apps a lot of people will talk about our bare metal cloud some people call the infrastructure our bare metal cloud actually that's not quite accurate either bare metal is a key component of OCI Gen 2 OCI but it is not all of a Gen 2 OCI we have bare metal compute servers with the classic two socket Intel machine we have dedicated X to the data machines we have a lot of stuff in in our Gen 2 cloud in our Gen 2 infrastructure bare metal is just a key and important piece all right so the primary reason we we architected our cloud from the ground up was security security built in from from the center to the outside from the perimeter to the inside impenetrable barriers autonomous robots but there are other reasons we built Gen 2 of the cloud as long as we're rebuilding our cloud as long as we're we architecting our cloud there were other design goals and another design goal is we want to make it very very easy for you to move your existing data center your existing workloads into the cloud NS ap application running on an Oracle database just lift it up intact with all of its customizations and everything drop it down in our cloud ebusiness suite any any any bespoke any custom application you have especially an Oracle database based application lift it up move it over no fuss no muss press about very very easy to do that's not what most clouds do that's not what they were designed to do ten years ago we want to make it very easy our goal has always been as we move from one generation of computing to the next as we move from mainframes to many computers to client-server pc networks to internet architecture to the cloud every generation we want to protect your investment in data in applications and all that code and make it easy for you to lift that stuff up move to the next generation and take full advantage of the next generation we did that with the gen 2 cloud and the gen 2 cloud add a lot of automation so save you save labor costs eliminate errors through automation and the overall price performance should be much better is much better than you get on-premise ok other clouds been around for a long time they were not really designed for the enterprise security is not good at the current the current state of the art and believe it or not if you move an on-premise application into the cloud it will cost you more not less to run typically all right the Oracle generation 2 cloud again the design goal one platform one secure platform to run everything all your applications the Oracle dawn autonomous database was built on top of the gen 2 infrastructure our new South all the fusion SAS applications today are already running on not every customer undone SAS but we've already got the fusion applications and the NetSuite applications running on the gen 2 cloud infrastructure it's designed to run all of your existing enterprise applications lift them up move them over but also cloud native applications using the coolest new cloud technologies docker containers kubernetes will your favorite open source stack bring we run it it's probably there you don't even bring it it's got all that stuff waiting for you so you can run you can build cloud native applications in the same environment Argenta our Gen 2 infrastructure that you run your enterprise applications Gen 2 is available now in the public cloud and is available next year with clouded customer oops back it up one slide I'm sorry and if you got if you got database cloudy customer this is very important I think if you are currently having you know bought database clouded customer come this summer come this next summer you can press a button and we'll upgrade you free upgrade automatic upgrade from gen one database cloud a customer to Gen 2 autonomous database cloud a customer so with Gen 2 we can actually deliver our autonomy that autonomous database to your premises in the form of database cloud at customer Gen 2 so we're very excited about that and it's an easy upgrade for what's proving to be a very popular product our database cloud at customer all right other design goals again as you move to the cloud we want to provide better functionality and better performance than you have on premises all the latest tools Exadata rack data guard you know everything run give you a choice between bare metal compute and sharing sharing compute with with other customers would via VMs it's your choice much more automation upgrades are automated for you patches are automated for you press a button migrate a workload very important easy to you know easy to build a big pipe between your data center and our cloud and you'll see that you know we we don't charge you a lot of money for moving data back and forth in and out of our cloud and we guarantee performance worth because we don't over oversubscribe our infrastructure we can guarantee we'll give you an SLA we'll give you an SLA that guarantee your performance nobody else does that all right two things are true about Gentoo cloud infrastructure our Gen 2 cloud infrastructure it's faster than any other cloud it should be it's newer it's brand new and we we priced it very aggressively the combination of being much faster and being aggressively priced is good for you because rich is much cheaper than the competition all right we'll compare ourself to Amazon and here just plain old raw compute were 50% almost 50% faster in compute five times faster in blocks or just just performance not cost the next slide is cost causes costs even better so we're 50% faster and compute five times faster and block storage and twice as fast networking but when you compare price amazon's basic compute is triple our basic compute when you move to our cloud you do save money you move to their cloud actually cost you more their block storage cost 30 times what our block storage costs and their networking again you know you move data out of your out of your data out of the Amazon Cloud bad idea because it costs more to move your data out of the Amazon Cloud then it costs you to move your data out of the Oracle cloud we believe you should be able to move data back and forth amazon's pricing is very clear move data in your done block storage you know block storage six times faster but 97% cheaper you know so performance advantage significant six price advantage much more significant big data workloads this is a terasort workload by the way it's a big data workload we're 3x faster but again eighty percent more than 80% cheaper huge huge cost differences the cost differences are much bigger than the performance differences the cost savings is much more than the performance advantage because it's a combination our cost savings a combination of being faster and being priced more aggressively we have this new RDMA network that's remote direct memory access it's really ability of computers that are networked together to move data from one computer directly into the other computer without taking an operating system hit without without an interrupt to just get you know goes direct from one spot to the other it creates a very low latency network you can move data very very fast it's a fact by the ways how Exadata works Exadata has an RD i mean the way the exited the machines are connected the reason exited is one of the reasons why exited is so fast is because it has an hour DMA network and now we've used the our DMA network in lots of areas of our cloud to cluster machines together gives you one and a half microsecond latency very low latency and a huge bandwidth 100 gig gig bandwidth and this is a computational fluid that's it's a scientific workload an engineering simulation CFD computational fluid dynamics bunch of partial differential equations doesn't really matter what you know how what it is the fact is we run this four times faster and much cheaper you know much cheaper than Amazon same you know more examples of CFD is computational fluid dynamics more simulations we just run faster big performance advantage much bigger price advantage compared to Amazon okay uh our Gen 2 clouds security security security Star Wars cyber defenses so we've used a lot of the latest artificial intelligence and machine learning technologies to build these robots to protect the cloud infrastructure to find threats and kill them to search and destroy our robots versus their robots you're not fighting only with what both your hands tied behind your back anymore we protect against a distributed denial-of-service attacks we have a Web Application Firewall to look at messages coming you know coming in - inspecting the messages - coming into wet web applications the autonomous database we've already already talked about obviously we we manage all of our resources we keep track of you the we keep track of users then we keep track of what resources users are allowed to look at the autonomous database automatically discovers threats automatically repair threats while running everything's encrypted of course at rest and in transit we have a key management system where you can manage the keys to your own data it's backed up again it's got a hardware security module that allows you to manage the entire lifecycle of your encryption keys this is again a very big deal you don't have to trust us after all I just said earlier we don't trust you not all of you so you don't need to trust us that's very important you want to manage you want to manage your encryption keys god bless you you do it we provide all the tools you know to help you do it in an efficient and a secure way the cloud controlled computers the new architecture these impenetrable barrier is protecting the cloud perimeter and every user from every other user full physical isolation from other tenants is built-in and that's why we architected our cloud I mentioned the key management service lets you manage your own keys of course we have a Casspi cloud access security broker we're constantly looking around for threats you know we started with you know with the autonomous database you know trying to you know identify threats now we're looking for threats throughout the cloud with all of the assets that we're managing search autonomous robots Search and Destroy threats we check for configuration changes anomalous configuration changes people people go here and that's what they do when people with threats they start changing changing your configuration they make all of these alterations we look for those things and if they if someone you know maliciously changes the configuration we discover that restore the configuration kill the threat web application firewall protects your infrastructure based applications and our denial of service protections guarantee you bandwidth even with a huge attack designed to manage manage through the worst you know layer 3 and layer of layer 4 attacks and still make sure your your service is running ok in 2019 this is what our our our footprint looks like globally with our generation to cloud but that's not every location we're supporting in fact we just added another location I think agreed to add something in UAE I think yesterday or the day before so we're continuously adding to this but on top of that we're willing to build additional locations for our gen 2 cloud we're built willing to build our gen 2 cloud on your data center floor just for you no caveats identical to what we have in our public cloud with the barrier with the cloud control computers all the way all the autonomous robots everything we will put in your data center just for you for our most security-conscious customers for our customers who doesn't want to have the broadest bandwidth between their data center and our cloud will put the cloud will co-locate them we will co-locate them okay when you come to Oracle and you buy infrastructure and database services you get our gen 2 cloud now that's the only thing we are selling across the cloud we are dedicated to this new secure cloud it's the only thing we are selling as I said everything is going in that direction all of our SAS applications all of our customers we think this is what we should be doing this is what you get our clouded customer which is Gen 1 for a few 4 months don't worry I we think that's a terrific product and come summer press a button and you get all the autonomous services without we sell although all the appropriate hardware when we deliver the gen one cloud a database cloud customer we delivered the Exadata is all the appropriate gear so when we have the gen 2 stuff ready you know for cloud a customer you push your button upgrade that software you get it automatically and I think that's and that's been our database cloud to customers been one of them our most popular products and we make it easy for all of those customers to upgrade and then as I just mentioned full OCI everything bare metal you name it everything load balancers blahblah everything Donna mr. booth you can get that and also in in 1900 k that's our Gen 2 cloud the most important component perhaps of the gen 2 cloud is our autonomous database and we've made a lot of progress over the last 12 months since we gathered together here last year but the autonomous database in it you know we've announced last year we announced for data warehousing and then a few months ago we announced the Dinamo's database now also handles transaction processing so it really handles all of your workloads and when I mean autonomous I mean autonomous I read an article by the way to folk Amazon did not write the article someone who's writing the articles saying well Oracle may have an autonomous database but Amazon is developing these semi autonomous databases semi autonomous databases like like a semi self-driving car you get in you drive and you die it's I didn't know what it is I mean I you know okay whatever this thing is fully automated fully self driving it provisions itself it gets servers to get it gets IO it gets network capacity if if the load grows it gets it automatically scales itself it automatically creates and drops indexes all the security is automatic if there's a threat that discovered the threat is killed automatically while the database is still is still running if if the server fails you know database server fails the database keeps running if the database software fails the database keeps running it autumn he backs up the self it automatically recovers itself etc it's fully automated it really is self driving no one's going to die and back it up I'm sorry not back once I because there's nothing to learn and nothing to do that makes it really easy to use here's the manual now here's the list of things to do we're finished training course over okay and it's also a very low cost to operate because there's no Associated labor but also as you as we'll see in these benchmarks just the running costs are much lower than a conventional database all right the Oracle autonomous database is serverless and true truely elastic what do I mean by service I mean when the database isn't running when you're not running the database you're not paying for any servers when you're not running the database you have zero servers allocated to you so you pay nothing for servers you still pay for storage you don't want us to throw your Daito away you want to keep the data see a paper storage but if you're using the shared server version of the autonomous database and we have two versions when you're not running it is service it is a true case it's a true case of you only pay for what you use while you're using it it's very different than other clouds which says I wonder you know I want these four servers for redshift I want eight sir eight eight cores for Aurora and you own those whether it's running or not it's not a serval or service this is truly truly serverless and adds capacity is it needs capacity for a month and close and then it releases all that capacity truly elastic and automatic this is one of the reasons we can guarantee if you use our database versus the database any database at Amazon we guarantee we guarantee bringing your Amazon bill we guarantee to cut that bill in half maybe the most famous aspect of the autonomous database automatic discovery of threats automatic remediation of threats while the database is still running the database basically never stops running I you can upgrade it to new versions of the software you can patch it hardware can fail software can fail none of it brings the database down all of those things bring the Amazon database down everything I mentioned brings the Amazon database table none of that if that's the Oracle database one of the reasons we can guarantee that the Oracle database is at least 25 times more reliable than the Amazon database 20 times 5 times more available than the Amazon database ok on again if we eliminate human labor we eliminate human error the great thing about auto pilots is they don't make mistakes they make the or they make the same mistake over and over again why you pick it then that's a bug but the autopilot no human labor no human error huge improvement in reliability of your systems much lower cost in terms of maintaining those systems Amazon fails again for hardware failures software failures backup upgrades we don't we keep running next slide alright um by moving the autonomous database does that mean that all of your data you know your database administrators are going to be out of work not hardly there is a an incredible shortage of skilled IT professionals on the planet Earth and it's good if we take away some of the mundane drudgery about running a database so they can use their talents to bring up new systems use their talents to do a better job on data analytics rather than making sure that the database is backed up once every day then scheduling those manually schedule those back up once they really are having to physically recover the database if it ever goes down all that stuff is automated your developers your administrators are you know now working on tasks that are a higher value for the business the the database cause you know automation is great because typically people build a data warehouse let's say and they tune it and then six months later the data warehouse is five times ten times bigger than it was when it came online and the tuning that was done six months ago is obsolete the Oracle autonomous database constantly Tunes itself and retuns itself and retuns itself automation automation is great so something that human beings just can't can't shouldn't do the system constantly adapts the changing shapes of data to changing workloads constantly retuns and optimizes itself every time it runs we've been working on this for a very long time we've been optimizing and automating our database software for years and years and years this isn't something we just came to all of a sudden we've been adding automation at automation adding automation adding automation with every generation until we automated just about everything and then we became autonomous but it wasn't enough just to automate the database software itself to be autonomous the you know when you're running a database you're not only running the database software you're running the underlying operating system and hardware and servers and network and there are all these other things you have to keep track of you have to add storage capacity on all of that we've also been automating database infrastructure for years in the form of Exadata we've been automating the network and optimizing the network and automating storage and backups and recovery and all of these things we've been automating it automating things we've been optimizing the operating system and the VMS that run the database we've been doing that for years and years and years and finally when you combined automated infrastructure with automated database with automated data center operations and you completely automate those three parts you get the world's first self-driving autonomous database it took a long time to do okay provisioning it runs on that say data or the autonomous database runs on Exadata it uses RAC real application clusters for scale out and it uses active Data Guard to create a copy a copy of the database in a remote location so it's highly available if you know if the server fails it keeps running if the data center fails the active Data Guard version of the of the database it keeps running so you've got complete disaster recovery built into the autonomous database security several things on that just a few new things one of the interesting things about the Oracle database is the the database administrator who created the database and can see the structure of the database cannot see user data it's the only database where the administrators can't look at user data we're the only ones that do that and it's a feature called data vault the management you know management patches not just the database but also the operating system online we have a version of the Linux operating system that our database runs on the UH next data that we run on exa data and we can patch that Linux operating system while the system is still running let's see when a server fails when database software fails and one you know one of our rack nodes goes down the application doesn't see it there's an application transparent failover it's not that it automatically recovers itself and then a bunch of users are kind of kicked off the system there were on that node that's not what happens the users never see the failure there might be a slight delay in processing maybe they'll see a one second two second delay but the application keeps running the users keep running it's completely transparent to the users data center fails your data your database fails over to a new location application keeps running uterus keep running it's serverless you only pay for what you use and it constantly optimizes itself using machine learning and we get better and better at this every year our optimizations are more efficient for example the latest version of the database never has a query performance regression in other words you know typically we come up with a new version of the database we've improved our optimization and out of a hundred queries you've got running 99 of the hundred run faster but one of them runs slower that's just the way it would work for years and years and years of Oracle new version we made most of the queries run faster with our new optimizer but there are a few outliers casas where the query ran slower that never happens anymore because now we put the new optimizer in it recompiles all the queries runs the you know it runs the query then then you know and and in the new mode to say is this faster that is the new mode faster or the old mode faster if the old mode the old optimized queries faster than the new one it will use the old optimized query not the new one so 99 of your queries will run using the new optimizer and one query will stay on the old optimizer to avoid all regressions by the way customers tested the hell out of our database to find these query regressions it was a lot of labor on your part I can't adopt the new version of Oracle because you know I got to make sure that you know some things don't you know what if some things run slower nothing runs slower anymore autonomous database for data warehousing autonomous database for transaction processing one is design for queries than analytics and the other is designed for transaction processing and a combination of transaction processing and reporting combination of mixed workloads and that's very interesting that we have one database we have one database for data warehousing Amazon has one database for data warehousing also called redshift we have one database for transaction processing by the way it's the same database we use for data warehousing Amazon has a totally different database to use for transaction processing called Aurora this is a huge problem for mixed workloads which I will demonstrate shortly okay um data warehouses we use a columnar format transaction processing we use a row format different kinds of optimizations and here's the data warehouse in action so these are the blue but the the blue bars are customer tuned version of this application the red bar is we just dropped all the indexes and re ran it with the automated data warehouse it turns out that the machine learning the computer optimized version of the application ran much faster than the human optimized version in every case for the stock exchange for the manufacturer for the bank the automation is very effective and the automation again it keeps automating over and over and over again so a DW works extremely well even when compared to experts who have tuned human experts who have tuned this thing you know took a long time before a computer could beat a human being at chess but that happened a while ago and the latest technologies the latest technologies computers are getting pretty good using machine learning in terms of doing certain kinds of complicated tasks that human beings do and computers can be very competitive to take over those tasks and this is one example where computers are now better than people even something that is professionally tuned like the NetSuite workload NetSuite you know a lot of smart people working on it for years and years and years you know the guy who did did a lot of the before London is line the performance was first in his class at Harvard and mathematics big deal still wasn't as good as the computer you know and he's pretty smart and there's a lot of very talented people but even in this case worked years and years and years of experts tuning an application the machine learning the machine educated robots were better at tuning the application than the smartest people and they used a lot fewer indexes a lot fewer indexes than than the human experts used an announcement the early versions of the autonomous database were on a shared Exadata on a shared Exadata machine it was a server list design in other words in other words we allocated a server for you when you ran and then when you stopped running the server went away and you didn't pay for it so it was a server list design to give the very lowest cost provide the very lowest cost for our autonomous database users we still have that option we still have an option to run shared service however if you're very security-conscious and you don't want any other users in your Exadata machine you don't want to share a DEXA data machine you want dedicated infrastructure you want complete isolation from other customers we will we will provide you in our public cloud 100% isolation your you are the only customer you are the only customer on that X data machine and we've got impenetrable barrier around that machine so that no other customers can get to it it's completely isolated we use our cloud control computers to surround that machine and nobody can get at it same technology we used we develop when we architected the cloud it really is a unique feature of the Oracle generation 2 cloud so you completely control that machine okay autonomous database cloud a customer I think I've actually mentioned this already that right now we're on version one of our autonomous database cloud a customer but this summer press a button and database cloud at customer becomes autonomous database cloud a customer full autonomous capabilities press one button again this is one of our most popular offerings and it's going to be upgraded in place at your data center to have full autonomous database capabilities in your data center with a broadband link a broadband pipe to all of your applications in your data center you don't have to buy a new license to use our cloud you bring your existing database license your existing middleware licenses you come to the come to the cloud you get exadata infrastructure autonomous database every everything is automated you pick shared server lists and you get really really low costs guaranteed to be half the price half the price of what Amazon charges you guaranteed more importantly there's no labor there's no human labor there's no human error there's no downtime your developers are more productive they get new projects to market the analytic applications faster lots and lots of benefits besides so if you want all of those benefits you've got to be willing to pay less all right this slide says we run five to ten times faster than Amazon simply not true I think the best benchmark we have here unless cut you count the one that's infinitely faster than Amazon is 80 times faster than Amazon okay so I'm gonna go and I'm gonna run a serious okay let me just look at I think got one more slide here again continuous perfect you know three to a hundred times faster than Amazon it really should be three to 80 to infinite depending on what you're doing automatic tuning runs on that's the data fastest infrastructure tuned by computers and machine learning delivers the lowest cost highest reliability highest availability I'm going to just run a few benchmarks that take me about five minutes and then we're done for this morning or this afternoon okay so here is the first he's gonna be interesting now we ran this benchmark last year and I wanted to make sure that we were still had advantage over Amazon redshift so we ran a whole series of Amazon redshift date benchmarks and this is a member of redshift is there a data warehouse so this is the autonomous data warehouse running against against Amazon's redshift okay stop okay so word we're done they're not done and so they got a long way to go we're about eight times I believe I believe we're about eight times faster still we're eight times faster last year we're still eight times faster all right they go to here they said my problem I wasn't pressing hard enough on this menu I've never heard that explanation before there we go all right all right so this this is so that was the data warehouse we're still eight times faster than Amazon's data warehouse this is autonomous transaction processing not the data warehouse versus Amazon Aurora Aurora is their specialized transaction processing system and let's just start this one up I know last year we let Amazon start first but I'm not going to do that this year no more mr. nice guy we're gonna start at the same time and let's see how this goes okay we're ten percent complete there they're getting close there one percent complete doesn't look good for Amazon but they're trying I wish we'd go faster we're 20% they're still at 1% oh I got 2% dammit okay you know just rough estimate were 10 times faster eight times faster here also but these aren't the interesting ones but again about eight times faster and much cheaper all of a summary for all of this let me just stop this stop stop okay no I'll do this and now this is my favorite remember I said you know Amazon has a data warehouse database redshift and they have a transaction processing database Aurora well what if your application has transaction processing and reporting like that's a really bad idea you should never ever do that because when you have two we have one database that does both things but they don't do that so what happens when you have this mixed workload of transaction processing and reporting again Oracle has one database that does all sorts of things we do handle IOT we had a real time we handle XML we handle all sorts of interesting things I mean the basics of can we do reporting and transaction processing with the same database yes we can we've been able to do that forever and the son still can't do that they're about no no no 10 20 years behind in database technology not even their database technology aurora's my sequel they just took open-source my sequel which we develop renamed it made it no longer open-source you know they call it Arora they took redshift from somebody else and just never it but they didn't develop any of this stuff let's not their business okay so I'm gonna stop this but we're 49% they haven't gotten to 1% yet we're and we're about I think 80 times faster than they are in a mixed workload by the way most workloads are mixed workloads that's a problem at Amazon ok now let me show you Amazon's best database by far it's not Aurora it's not redshift it's Oracle what happens if you run Oracle on Amazon well it's much better but it's still not as good as our cloud we're still much faster you know I'm worth about 3 times faster and about half the cost for a little more than half the cost when you run Oracle on our cloud versus Oracle on their cloud but that's not the reason you run or a claw on our cloud is opposed to Oracle on their cloud it's not that we're three times faster and less than half the cost that isn't the reason I can't let this go a little bit further but I've got a summary that will show again we're about three times faster and less than half the cost let me stop this and show you the next benchmark the last benchmark ops actually this is really cool because this is the Oracle autonomous database running in our gen 2 cloud versus the Oracle database running at Amazon and what we're going to show you is what happens when our autonomous robots find a security threat they immediately spring into action and patch the threat remediate the threat kill the threat destroy the threat now how does that compare to what goes on in it what happens in Amazon ok I'm good back here all right yeah ah doesn't look good I know if you're looking at the top bar this top thing right here and Amazon's having a problem they are now patching the threat RDS that's the Amazon Oracle service it's down system you have to take the system down so in this case we're infinitely faster than they are you have to shut your application off in fact a human being has to figure out to shut the application off shut the application off patch the application manually bring it back up systems not available by the way with RDS you're still paying for it it's not server less evil in the system isn't running you're still paying so which is kind of painful so it's down but you still pay for it I'm gonna stop it then show you the summaries okay so okay these are these are all the summaries so the first thing we're still okay we're not we're nine times faster eight times cheaper on a day would data warehouse application in the first OLTP application against Aurora we're 11 times faster and eight times cheaper these are all real benchmarks they're running outside you can check them this is the best Amazon has to offer this is their latest greatest best stuff okay in a mixed workload which is most workloads a combination of OLTP with some reporting we're a hundred times faster and eighty times cheaper that's a big advantage eighty times cheaper is good you can go to Amazon if you want to but you know it's very expensive a lot of people just think wow the Amazon Cloud really cheap not true not true Oh LTP okay the what's the best database you can run on Amazon Oracle but even when you run Oracle on Amazon it's gonna be you know we're gonna be three times faster and less than half the cost and you get and Amazon has none of the autonomous features they have none of the reliability features they don't have rack they don't have self tuning they have no autonomous features not available they don't have active Data Guard they have no disaster recovery they have no server failure recovery they have no software failure recovery they have no automatic patching you know they've got none of that they've got none of that and we automatically patch who you saw we keep running the system keeps running in this case we are infinitely faster and infinitely cheaper but just take the simple mix workload yeah eighty eighty times cheaper is good I mean that's a that's a good competitive advantage you'll never see that in the Amazon Amazon retail store you'll never see them selling you like a polo shirt that's eighty times more expensive than one of their competitors okay we invite all of our customers to try Amazon polo shirts and our autonomous database the okay and we've we now have free trials out there we think try it I think you'll like it in summary we've really are connected our cloud from the ground up to develop these new Star Wars cyber defensives impenetrable barriers protect the perimeter of our cloud and keep threats from getting in and then barriers and pendel barriers around each customer zone protect threats from spreading laterally from one customer to another we have these separate separate negative cloud control computers that are virtually unhackable because we don't share them with anybody there's no user code in our cloud control computers a fundamental architectural difference between the way we built our cloud and the way Amazon built their cloud throughout our cloud we have autonomous robots searching for threats not just threats to our database which automatically repairs itself while running but threats to any aspect of our cloud we search and destroy those threats it's our robots versus their robots it's our robots that keeps your data safe and prevents people from stealing it and you finding yourself on the front page of a newspaper above the fold there's no delay for human processes there's no time for human processes we have to react automatically we have to react autonomously to these cyber attacks the autonomous database you eliminate human labor you eliminate here there's nothing to learn nothing to do your developers become more productive they bring up new applications they did a better job of analyzing data your system is more reliable it never goes down it never goes down thank you very much [Applause] you
Info
Channel: Oracle
Views: 107,231
Rating: undefined out of 5
Keywords: Larry, 2018, second cloud generation, Keynote, autonomous services, autonomous future, Oracle OpenWorld, retire-2018, cloud computing, enterprise workloads, enterprise cloud, Keynote Presentation, OOW, Oracle Open World, 30726466, cloud at customer, enterprise data, Autonomous, CTO, LJE, Cheif Technology Officer, Executive Chairman, Keynotes and Presentations, data security, public cloud, cloud generation, Autonomous Database, Ellison, Executive Chairman and Chief Technology Officer
Id: b5qZVk0F_yg
Channel Id: undefined
Length: 66min 55sec (4015 seconds)
Published: Tue Oct 23 2018
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.