Oracle Live: Oracle Database Breakthrough Innovations
ANDY MENDELSOHN: Hi, this is Andy Mendlesohn, EVP of Database Server Technologies at Oracle, and I'm going to talk to you today about some breakthrough innovations we've been coming out with with the Oracle Database. OK, so with Oracle Database, we're working on three main focus areas-- working on core Database software, working on the optimized platform for the Oracle Database, the Oracle Exadata, and we're talking about Database Cloud Services. And so today, we're going to focus in on the top two areas, Oracle Database software and Database Cloud Services. For database software, the big space we're working on is something we're calling converged database, and we'll talk about that more on the next slide. And, of course, we're continuing to work in areas like high performance and scalability and reliability, as always. Oracle Exadata is evolving, and we don't have big announcements today, but look forward to big announcements coming later this year.
And for Database Cloud Services, our big story is around Autonomous Database. It's a radically new way of providing database technology really easily, and easily consumable by any developer, by any DBA, by any analyst. Everything is automated.
And uniquely, versus other cloud services, our Autonomous Database Cloud Service is available both in public cloud and also in customer data centers via something we call Cloud@Customer. Let me first explain to you what we mean by the term converged database. So you all have your smartphones. And you remember 10 years ago, before smartphones, you used to have a whole set of different devices-- dumb phones. They did messaging, they took pictures and videos, et cetera.
And the smartphone won because it converged all those different devices into one device that we now know and love called the smartphone. And it won because it's much more productive for you as a consumer to use that one device. The notion of converged database is exactly the same.
We are converging lots of different database technologies, that heretofore had been in separate databases, now into one, single database we call a converged database. And the reason it's going to win is for the same reason-- it's going to make developers and analysts much more productive to have just one database that lets them manage all their data, any workload, and just makes them in general much more productive because they can use industry standard SQL against all their data with all their workloads-- the same dialect of SQL everywhere. And it's always good to compare what we're doing with what the other guys are doing. And Amazon and a bunch of other niche vendors are going down this other path, where they're saying, you know, you can't have the best database for a given workload if it's trying to do everything at once. You need a specialized database.
And so in Amazon, they have over a dozen different database services for doing all kinds of different workloads. They have Aurora for their transactional workloads, and they have Redshift for analytics, and they have Dynamo DB and JSON document stores, et cetera, et cetera. And what we are doing is basically saying, we can do all those workloads better than them in one place, and so if you're a developer, why would you want to use five different databases to have to build your application when you can just do one? You have five databases, you have to fragment your data across those databases. It's impossible to run transactions against them, it's impossible to do disaster recovery. Security is all screwed up because now security is the security of the least secure of your five databases, et cetera, et cetera.
So our strategy is give developers and analysts one converged database that has all their data. You could use open standard SQL for building all your applications against that database-- the same SQL dialect for all your different workloads-- and go from there. So this is a much more productive environment, and this will, of course, win the day. So let's now talk briefly about what we're doing in the cloud. Our big innovation there is something we call Autonomous Database.
And we came out with this product a few years ago, and I think you'll see later that pretty much everybody agrees this is by far the best cloud database out there in the market. What's really great about it is you get all the power of the Oracle Database that our customers know and love, enterprise-class database running on the most powerful platform for database, the Exadata Database machine. And it's now become fully automated. Everything is automated. You push a button, you prevision the database in a couple of minutes, it does its own patching, it does its own tuning. We call this all self-driving, self-scaling, et cetera.
And it offers true cloud elasticity. The other guys still are provisioning shapes to run their databases, and a shape is really like a physical server. We are provisioning exactly what you want.
If you want 17 CPUs to run your workload, you get 17 CPUs. You don't have to pay for another 32 just because that's the next biggest shape available. And when your workload has spikes or has bigger processing requirements at the end of the month, you don't have to provision a big shape for that because we auto-scale for you. So if you initially say you want 17 CPUs, but then the end-of-the-month processing needs 64, we will automatically scale up from the initial number you give us, up to 3x that, to get your workload done.
So that's true cloud elasticity. And finally, if you're a developer analyst, we have by far the most productive environment for building data-driven applications out there in the market. It's called Oracle APEX. It's integrated with Autonomous Database. And we'll talk a lot more about APEX later, but it's a low-code environment.
Let's you build applications 20 to 40 times faster than traditional coding methods. So although there's just one Autonomous Database architecture, we are providing three separate services that are optimized for three separate use cases-- so Autonomous Data Warehouse, Autonomous Transaction Processing, and Autonomous JSON Database. Autonomous Data Warehouse is our offering for any kind of analytic database use case-- everything from a data warehouse to a datamart to performing machine learning against your data, a data science laboratory, et cetera.
Autonomous Transaction Processing is for running your operational workloads, things like transaction processing, mixed work loads of transactions and analytics, and everything else in between. And then finally, Autonomous JSON Database is a specialized version of Autonomous Transaction Processing specialized just for the JSON developers who just want to build applications using a JSON document-style API. And, of course, because it's just full Autonomous Database, if and when you hit the wall with that JSON Database API and you want the full power of Oracle SQL, you just push a button and we can give you, in place, all the power of Autonomous Database without any movement, without rewriting your code, et cetera. So I'm going to quickly now go through some of the new offerings we've had in the cloud space over the last six months or so.
Number one, Autonomous JSON Database-- we talked about that already. So this cloud service is for developers who want to build JSON document applications. This is a space that was popularized by MongoDB. And what's really interesting about our offering here is that we offer an experience very much like you get with MongoDB, but we offer it at a lower price, and we give you better performance-- up to 2 times better performance than MongoDB and Amazon's DocumentDB. And you get all the benefits of Autonomous Database here.
It's all self-driving, enterprise-class, fully transactional, et cetera. So last September, we introduced Autonomous Database Cloud@Customer. So this is exactly the same Autonomous Database service that we first delivered on our public cloud, is now available to run in your data center.
The exact same experience. We prevision the database via the same web browser console. The only difference is now when you specify where the service should be provisioned, you can now specify, in addition to one of our public cloud locations, one of your data center locations.
The same business model is in place, et cetera. Very, very unique experience that we give you the exact same Autonomous Database experience from the public cloud in your data center, as well. Number 2, or number 3 actually, in December, we have enhanced our Data Safe service.
So Data Safe is something we came out with about a year ago, and it's a single console that lets you manage and oversee the security of your databases. The big thing we're announcing now is it's now multi-cloud. So originally, your databases had to be in Oracle's public cloud. Now the databases you manage can be anywhere. They can be on-premises, they can be on other public clouds, and so now you can have a single view of all your databases no matter where they are, across clouds and on-prem.
And, of course, Autonomous Database is very successful in the market, growing very fast. We have all sorts of customers, from the small startups to the big enterprises, running on Autonomous Database, doing things like datamarts, data warehouses, and transaction processing, et cetera, et cetera. And don't take our word for it that Autonomous Database is the best database cloud service out there in the market. Gartner, who is a very well-known analyst firm, just recently, in early December, produced their latest Oracle Database Cloud Critical Capabilities report where they analyzed different use cases-- four use cases for operational databases and four use cases around analytical databases. And across these eight use cases, Oracle was either number 1 or number 2 in every one.
We were number 1 in all four of the operational use cases-- things like transaction processing, mixed workloads of transactions and analytics, event processing, et cetera. And we were either number 1 or number 2 in all four analytical use cases-- things like data warehouses, operational analytics, et cetera. So, again, if you go up to oracle.com, we have links there to these reports. And one of the things you'll notice is that all those popular cloud services you hear a lot about on Amazon or on Microsoft, they're all ranked much below us on these Critical Capabilities scores. So we don't win by just a little in most cases, we win by a lot.
So take a look at that if you want to see what a third party analyst firm thinks about our products. And, of course, Oracle Database is very popular in general. The best gauge of how good your product is how widely deployed it is. 97% of the global Fortune 100 companies use Oracle Database. So now we're going to sort of segue onto what's coming next for Oracle Database on the cloud. And the big, new thing we're announcing today is what we call Oracle APEX Application Development Service.
So earlier, I talked about APEX a little bit, but let's go in a little more detail about what APEX is. So number 1, it is a low-code development environment and deployment environment for data-driven applications or database-intensive workloads. It's a feature of every Oracle Database running on-prem today, every Oracle Database on a cloud. It's there.
And it is a very interesting environment. It's something that lets you declaratively build your applications without coding, and it's something we've been working on now for well over 10 years. It's a very mature, low-code tool. It's not like I know a lot of people are now coming out to market with low-code tools. This one is really mature, really powerful.
You can build applications 20 to 40 times faster than traditional coding. 20 to 40 times faster-- that's really big. And on our website, we have a time and motion study that we did with an analyst to actually work through building a real application with traditional coding methods using ReactJS, and compared that to the time it took to build the same application in Oracle APEX, to just prove that what we're saying here, 20 to 40 times more productive, is really true.
And one of the big things-- you know, this tool is really mature. And what we have found now is that virtually any application you want to build in your organization can now be built with APEX. There's no need to use traditional coding.
In fact, traditional coding should now be the exception, not the rule, about how you build applications. This is a really, really powerful environment. And finally, we have a huge, huge community of users with APEX, and any kind of developer can use it.
You don't have to be a professional developer anymore to build powerful applications using APEX. You can be what we call a citizen developer, somebody who part-time is doing some coding, a line of business. You can be a data analyst who wants to analyze their data, but again, doesn't want to write a lot of code. This is a great environment for you to build your applications. And here, I just wanted to show you some screenshots of what kind of beautiful, modern applications, web-based applications, you can build using APEX. Because APEX is a declarative environment, it can very easily map what you want to do to any kind of device.
So if you have a big screen on your desktop, smaller screen on your iPad, tablet device, or even smaller screen on your phone, it will automatically adjust the visualization of your application appropriately so it's beautiful in all environments, and we wanted to show that here in some of these screenshots. And Oracle APEX is the world's most popular enterprise-class, low-code environment. We have over 500,000 developers using APEX today.
They are cranking out over 6,000 new apps every day. There are millions of applications in existence out there already. So again, if you go with APEX, you go with a really mature, really proven environment for building applications with low code really fast.
Traditional coding is a thing of the past. And what we're announcing now is something we call APEX application development service, and so what we do is we take Oracle APEX, and we make it the center of the world, and we deploy it on our cloud, both public cloud or Cloud@Customer, on top of an Oracle Database, the Autonomous Database. And, of course, the Autonomous Database runs on our powerful Exadata infrastructure, so you get a really easy-to-use environment fully managed by Oracle, both database-tier and middle-tier. And it runs as a service on our cloud.
And the best thing about this is it's also actually really low-cost. So starting at just $360 a month, you get this fully managed APEX service. Runs the mission-critical, enterprise-class database in our Autonomous Database.
Everything's fully managed. It inherits all the attributes of Autonomous Database. It's fully elastic. It can scale from 1 CPU up to 128. You can start with 1 terabyte and go all the way up to 500.
So it can be used for any kind of workload, small departmental workloads that just have a few users, and you can scale up to huge workloads with millions of users. So this is a really, really exciting offering, really nice service that nobody else can match on any other cloud, and you can try it for free today on the Oracle free tier. And now we're going to have Shakeeb Rahman, one of our architects for APEX, give you a demo of the power of building APEX applications.
[VIDEO PLAYBACK] SHAKEEB RAHMAN: Hi. I'm Shakeeb, and I build apps using Oracle APEX. APEX is a low-code platform that enables you to build applications fast, and it's used by customers around the globe, across all industries. One of my recent projects was focused on building apps that handle logistics for large-scale vaccinations, including scheduling, safety monitoring, and on-site queuing and processing. Now, these are mission-critical systems that need to be built fast.
In the real world, requirements change, and they change often. This is where APEX shines. Making changes to your apps can be as simple as dragging and dropping components on a page, and when you're finished, simply save and run your page again. APEX includes a number of powerful components, like faceted search that enables you to explore your data like never before.
And if you're not an expert in HTML or CSS, or you don't have an affinity for UI design, APEX has got you covered there, too. With features like theme roller and template options, you can easily customize and look and feel of your application in real time, without writing a single line of CSS. Everything you build is responsive, so it works on any tablet or mobile device out-of-the-box.
If you're familiar with SQL, you're going to feel right at home. With APEX, every feature of the Oracle Database is at your fingertips, meaning you can easily add things like spatial or analytic functions in no time. There's a whole lot more to APEX than what I covered in this short intro. To learn more, visit apex.oracle.com, and get started today.
[END PLAYBACK] ANDY MENDELSOHN: OK, so let's close on what we're doing next with the Oracle Database. And today we're announcing Oracle Database 21c. We're doing new innovations focused on developers and architects, analysts and data scientists, and also our traditional DBA ops users, as well. So before getting into 21c, let me briefly talk about 19c.
So 19c's our current long-term support release that's supported all the way to 2027. I know many of you are now already on 19c or in the process of upgrading hundreds of thousands of databases in 19c. You should all continue that process. So what is the objective of the new Database 21c release? Well, 21c is what we call an innovation release.
That means it's a vehicle for us to get functionality out to market really fast, especially to the hypercompetitive cloud market. The release is suitable for production use, but keep in mind that we will only support this release for two years. It's also a great release to use for building new applications that want to leverage the new innovations in the release.
Secondly, 21c is also a great opportunity for us in development to improve the quality of 19c. As we tested the 21c features, we often will find bugs that apply not just to 21c, but also to 19c. And so we will aggressively backport those fixes to 19c, thereby improve the quality of 19c, as well.
Also, a few of the more exciting innovations in 21c will be backported in 19c. As we come across those later in the presentation, I'll point them out to you. OK, so now I'm going to talk to you about some of the main innovations in Database 21c.
So first, let's talk about the innovations in 21c for the developers and architects out there who are building applications against the Oracle Database. First thing we're going to talk about is blockchain tables. This is one of my favorite features in 21c.
You go to Amazon, and Amazon said, yeah, blockchain is really cool, so let's create a whole new database service specialized just for blockchains. And if you want to do data management using blockchain technology, then you have to use this service. You can't use Aurora and Redshift.
And we thought that's a really dumb idea. If a developer wants to make their tables immutable using blockchain technology, let's just create a new feature in the Oracle Database so that when you create a SQL table, you can just say, this table is special, it's a blockchain table. I want it to be immutable, an immutable ledger table, so that I can query it using standard SQL, I can insert rows into it using standard SQL. But I can't update it, I can't delete rows from it because it's a blockchain immutable table. And so that's what we did in 21c, is a whole new example of converging what is otherwise considered a separate, specialty technology on Amazon, for example, into our Converged Database 21c.
Another area we've been working on for many years is SQL JSON. So we extended the power of SQL for query and updates against JSON documents many years ago. And in 21c, what we're doing is we're introducing a whole new representation for JSON documents that's much more efficient. We can now scan data up to 10 times faster. We can update JSON data up to 4 times faster. So all your SQL JSON applications continue to work in 21c-- they just work much, much faster.
So the next area I'm going to talk about are innovations we're doing for analysts and data scientists. So as part of our Converged Database technology, we added machine learning algorithms right into the database well over 15 years ago. So we've been incrementally adding new algorithms. In 21c, we've added some new algorithms for anomaly detection, regression, and classification analysis. And these algorithms can be used both for building new machine learning models, and actually deploying those models in the database against the data in your data warehouse, for example. The big, new innovation we're doing in 21c is we're adding AutoML on top of our machine learning algorithms.
AutoML is a breakthrough technology that now lets anybody-- a developer, a data scientist-- very easily create machine learning models just by giving us their data sets. Telling us, here are the features I want you to use to build your machine learning model, and here's the column that has the value I want you to predict. Oracle just figures out what is the best machine learning algorithm for creating a model around this data set, and it comes back and gives you some choices for you, and tells you the various levels of accuracy of the different models. Really, really important breakthrough technology in Database 21c for our developers and analysts. We've also continued to work on and improve our in-memory, columnar technology in the database. This is something we originally introduced way back in 12.1 of the database.
Very, very high-performance analytic technology. It's becoming much more self-managing in the 21c release. It's being used for more different kinds of SQL operations-- in this case, for different kinds of join technology. There's joins, as well, in 21c.
And we continue to improve the performance of our graph analytic capabilities in 21c, as well. And then finally, for our ops people, DBAs, people who care about the performance and reliability of the database, we've added a number of innervations there, as well. The biggest one is support for persistent memory. So let me just digress for a minute about what's persistent memory.
So Intel introduced a new kind of memory called Optane data center memory a few years ago. It's now available as a part of some of your Intel servers. And it's a very interesting technology because it's sort of a hybrid of DRAM technology and Flash.
It has high-performance read and write access almost as good as DRAM, but not quite, and it has the persistence capabilities of Flash storage in that you can actually use it as storage technology. So persistent memory from a database standpoint is a really, really, really important technology that we need to rearchitect our database technology to take advantage of. And 21c is the first release that we're doing this with. We've now added support so that if you want to, you can now store your data for your database in persistent memory, and use that as the storage for the database. You effectively have an in-memory database all the time. And this is one of those few technologies that we think is really important, so we are going to backport this support for persistent memory into the 19c release, as well, and you'll see that, again, some time during calendar year 2021.
Multitenant-- another really important innovation we first introduced in 12.1. It's now becoming widely used by customers who want to do either consolidation or private clouds on-prem. It's one of the key technologies underlying every Oracle Autonomous Database, every Fusion SaaS Database on our cloud, every NetSuite Database. And third parties like Salesforce are also looking at using this.
One of the big things customers wanted is they wanted the ability to have a standby database not just at the whole container database level, but at the pluggable database level. Pluggable databases are the virtual databases we offer you with your multitenant database, so in 21c, we've done that. We've added the ability now so if you have a problem just confined to one pluggable database, you can now fail over just that one pluggable database to the standby while you leave all the other ones where they are in place. Big, big, new feature. Plus we've added more and more powerful technologies for isolating your pluggable database, as well.
In the security space, a lot of you now have to file compliance regulations that require you to change passwords periodically, and it's sort of annoying because when you do that, you might break applications that use the old password. So we now have a way of gradually rolling over from the old password to the new password so that both passwords work for a brief period of time. So the last new innovation in 21c I want to talk about is better sharding. So sharding is a new capability we rolled out in database 12cR2 It lets you take a physical Oracle database and geographically distribute it across multiple shards or locations in the real world.
Of course, Oracle very uniquely has the ability to scale out, using our real application cluster technology, any kind of workload-- in particular, transactional workloads that no other vendor can deal with can be scaled out using real application clusters. Now, for those of you who have extreme scalability requirements or need to scale geographically, sharding is the answer. And we added that in 12cR2, and we have a number of customers using sharding for really extremely huge workloads that need extreme scalability. But the other use case we've been finding for sharding is something we call data sovereignty. So a lot of countries now have regulations that say information about the citizens of China shall be stored in a database in China, information about citizens of Russia have to be in a database in Russia. And so data sovereignty is a way of using sharding to satisfy these regulations.
So when you create a sharded database, it looks and feels like a single Oracle database, but with the sharding, you can say when rows about citizens of China or citizens of Russia are inserted, make sure you insert those rows into shards in data centers in those countries. And so we're finding this is a very popular new use case for sharding. So just in summary, 21c has a whole bunch of new innovations. We've done a lot of work to continue down the path of making sure we have the world's best converged database. We've been improving JSON data types, graph technologies.
SQL macros, again, make your SQL developers more productive. And we continue to work in the enterprise-class space, as well. And with that, I invite you all to go to our cloud today, and you can use Oracle Database 21c with the Autonomous Database free tier.
You can use it today with our database cloud services on the VMs and bare metal. And we will be rolling 21c out across all our other cloud services and our on-prem platforms Linux, Windows, exadatas, database appliances later in calendar year 2021. Thanks for coming. I welcome you all to come out to our public cloud and try out all the new innovations we talked about today-- Database 21c innovations, we talked about our new APEX application developer service, try that one out. And, of course, the multi-cloud version of Data Safe is available there to try out, as well.