I started working with SQL Server just as version 7 was released and I remember the consultants I worked with were very excited about this new version. It is almost comical to think about the tooling available at the time–especially around the concepts of security, availability, and performance. The management studio did not allow you to run queries–you had to open a separate application called query analyzer. As we begin migrating more customers to SQL Server 2019, I appreciate how far the platform has come and continue to be excited about the future.
To celebrate our 200th episode, we reached out to Bob Ward to come and celebrate this milestone with us. I have to plead the 5th on why it took us 200 episodes to invite him; however, we finally wised up and Bob walks us through some of the changes SQL Server has experienced over the years and what is in store (Spoiler—no big scoops) for the relational engine. We appreciate Bob joining us and look forward to another 200 episodes.
Bob Ward’s book: SQL Server 2019 Revealed
Constant Time Recovery
SQL Server on a Raspberry Pi!
Bob Ward is a Principal Architect for the Microsoft Azure Data SQL Server team, which owns the development for all SQL Server versions. Bob has worked for Microsoft for 26+ years on every version of SQL Server shipped from OS/2 1.1 to SQL Server 2019 including Azure.
Never Miss An Episode
Subscribe to get podcast notifications by email.
“If you’ve heard me speak about SQL ‘19, I don’t call it an engine anymore. Certainly, the engine’s a big part of it; I call it a platform, because it includes Big Data Clusters, PolyBase, machine learning.”
Read the full transcript of this episode.
Carlos: Compañeros! Welcome to another edition of The SQL Data Partners Podcast. I am Carlos Chacon. It’s good to be back with you, compañeros. We’ve been gone for a couple of weeks, so thanks for your patience there. We have an exciting episode, so Episode 200, it’s a little bit near and dear to my heart. So, when I started four years ago, I don’t know that I ever thought about what Episode 200 would look like, but here it is, and we have a special guest for you today. So, Bob Ward is in the house. Woohoo!
Bob: Happy to be here. Thanks for inviting me. 200? I can’t believe you invited me for your 200th episode. That’s amazing.
Carlos: Yes, well, you know, we had to get–
Bob: Very honored.
Carlos: Get the big guns out. Yeah, so thanks for being here with us. And so, we have changed up the program slightly. So compañeros, you’ll notice– of course, Bob you won’t know know, because you haven’t listened, but we have new music for the podcast, and so that hopefully is fun.
Kevin: So, we killed off Eugene, but then we brought him back immediately, because the fans demanded it.
Carlos: That’s right
Eugene: Yeah, well actually, no, I’m– what is it? I’m twin brother Boogene. We killed off Eugene and–
Kevin: V-gene, yes.
Carlos: Yeah, he gave his mother my cell phone number.
Eugene: I would not do that. My mom calls me like six times a day. I’ve actually got an Amazon Echo so that I can video call her.
Bob: Oh, my goodness.
Eugene: Well, it’s just been bad, because we used to take her into this Adult Day Care Medical Center kind of thing, and we haven’t been able to do that. I love my mom, but she’s bored and lonely.
Bob: I know. My mom is in the same situation right now. It’s really difficult.
Carlos: That’s tough, that’s tough.
Eugene: So, no, I would never give her your number. That would be a great unkindness.
Carlos: I’m just kidding. I’m just kidding. Okay, so we have Bob Ward and we’re going to kind of jump into it slightly here, a little bit, because his bio requires it.
Bob: It does? Oh no, what are you going to say from my bio? Dallas Cowboys fan, that’s what it should say.
Carlos: That’s right. Well, so I do have that and so one of the things that I should say is that people have suggested, “hey, you need to get Bob on the program.” And I’m like, “well–
Bob: Oh, that’s so nice.
Carlos: The problem is, he’s a Cowboys fan.”
Bob: Oh dear.
Kevin: That is a problem.
Bob: It goes against all your nature, doesn’t it?
Carlos: As a lowly, lowly Redskins fan–
Bob: That’s okay. It’s okay, it’s okay.
Carlos: This is the only time of the year that I get to enjoy, when they’re not playing.
Bob: Right. We’re all 0 and 0, right? Everybody’s the same.
Carlos: Okay, so for those of you who have been under a rock somewhere, I suppose, so Bob is the Principal Architect for the Microsoft Azure SQL Data Team. I hope that’s still correct.
Bob: That’s a long title. What does that even mean?
Carlos: But here’s the interesting piece. Now, you’ve worked at Microsoft and have been involved with every version of SQL Server shipped.
Bob: That’s true. That’s a true statement.
Carlos: OS/2 1.1
Bob: OS/2 1.1, yeah, that’s right.
Carlos: To the SQL Server 2019. Now–
Bob: This is tablet days, stone and tablet days, yes.
Carlos: That’s right.
Bob: Pretty much.
Carlos: I’m like, I don’t know that I have ever heard it referred to as OS/2 1.1.
Bob: So here we go, guys. Here, these are the diskettes. I’m not making this up. These are the diskettes to install SQL back then. So, I still have this in my drawer.
Eugene: Oh, you mean the Save icon? Yeah.
Bob: Yeah, pretty much. It’s true. 1993 is when I joined the company and OS/2 1.1 was what I first started working on back in those days. I know, it’s crazy.
Carlos: Wow. Yeah, so it’s come a long ways.
Bob: Oh, the evolution of SQL is amazing. I mean, think back to diskettes–
Bob: To Enterprise Systems with petabytes of data and huge scaled servers, now in the cloud, now in the Edge. I mean, you probably saw that announcement at Build, right? SQL on an ARMed Raspberry Pi device. So, SQL now on one of these, you know, a little Raspberry Pi. So, crazy. Great, amazing story, quite frankly, for the history of this product.
Carlos: Yeah, no doubt. Yeah, it has come in– I don’t know if different direction’s right, but so maybe simplistic to get it back on Raspberry Pi, but also the number of things that you can do with SQL Server now.
Bob: Well, the amazing thing, if you think about it, just real quick and make sure I threw in those little nuggets for you. I know you’re trying to get this started, here I’m just talking over you.
Carlos: No, no.
Bob: Hey, this thing right here is the same engine. I mean the engine that goes into this, it’s just a smaller footprint. It’s the same engine that powers ’19, it powers Containers, Linux, Azure. Everywhere SQL exists, it is the same codebase engine, we just have different flavors of it and different ways to deploy it, and sometimes we light up different features in certain ways. That’s one of the things that I love so much about working in this product, is even going back to the OS/2 1.1 days, I always worked with a group of developers that really believed in the core fundamental principles of security, availability and performance, and I call that the meat and potatoes of SQL Server. That’s kind of a Texan term, you know, meat and potatoes. But it is, it’s the core of what we do, and then we put all this stuff around it, so that’s what I love. So, this little thing right here, little Raspberry Pi is still got all that functionality built into the engine, so it’s pretty cool.
Carlos: Wow. That is cool. So, I hate to detract, but just a couple of things before we jump into the real, for real.
Bob: I didn’t mean to throw off your show. I’m never coming back, right?
Eugene: This is Carlos’s UserVoice tickets, right?
Bob: Yeah, exactly.
Kevin: My airing of grievances 1-40.
Bob: The airing of grievances, I like it. That’s right.
Carlos: Oh, so because Bob’s on the program and I saw that you have been involved with the SQL Server 2019 Revealed, we’re going to figure out a way– I don’t know how and Amanda is going to have a– her blood pressure is going to rise when she hears this on the recording, but we’re going to figure out a way to give away some books.
Bob: Well, I’ll tell you what I’ll do.
Carlos: Oh, here we go.
Bob: I’ll tell you what I’ll do. Figure it out. I don’t know how many are you thinking of, maybe, just off the top of your head?
Carlos: Oh, well so we’re a modest group, so single digits is what we were thinking. Maybe 10?
Bob: I will give you those books.
Carlos: Oh, there we go.
Bob: If you’ll give me your mailing address, I have in my office here some books, so whatever books you decide to do, I will donate those books to you, and I’ll sign them.
Carlos: Oh, well, so now, I was going to say the three of us might be on there. We might have to increase that.
Bob: Yeah, Bob, by the way, it’s going to be 10 books. It’s really 7 books actually. No, I’m really being serious. Happy to ship those books to you and I’ll sign each one of them.
Carlos: There you go. Compañeros, you heard it here. If you want a signed SQL Server Revealed 2019 from Apress Media, let us know. By the end of the show, I’ll figure out a way or I’ll think of a way for you guys to get that information to us. But as always, you can go to sqldatapartners.com and today’s actually keyword is going to be Bob Ward. So sqldatapartners.com/bobward or Episode 200 and we’ll have the information on there for you to grab those books. So, Bob’s actually done a little video, he’s kind of showed us how we’re going to start using that little bit. We’re going to try to go back to YouTube. I’m not promising demos just yet, we’re still kind of just audio, but at least you get to see us, which may or may not be a good thing. And so, if you’re interested in consuming it that way, that will be available here shortly.
Get your very own SQL Data Partners Podcast t-shirt!
So, t-shirts. We’ve talked about t-shirts in the past and we’ve made them available, but several of you have come up and said, “well hey, I’d love to support the podcast in some way. Don’t send me a t-shirt, let me buy it.” And so, we are going to make those available, so go to sqldatapartners.com/tshirt and that will kind of point you there. By the time this goes live, that will be available.
Carlos: And thanks, everybody for your support, there.
Kevin: Sqldatapartners.com/tshirt, singular?
Carlos: Oh, t-shirt, thank you.
Bob: That’s why Kevin’s on the show.
Kevin: I am here for the derail effect, actually.
Bob: It’s all coming together now.
Eugene: He’s here for like pedantry and derailing. So he brings us on task for tiny details and then way off task.
Kevin: I do.
Carlos: Okay, so with that, to jump into some of these pieces, and you mentioned in the beginning, so SQL Server has had this evolution. Now it’s on a Raspberry Pi, we’ve talked about Linux, so it seems to be, I don’t know, expanding in its scope. But at the same time, a lot of folks are nervous, and I’ll use the influence of NoSQL into the mix and particularly from a developer perspective it’s like, “hey, I don’t want to have to think about what my data looks like. I just want to be able to, you know, easy-button it. Maybe I’ll take some technical debt along the way, but I just want to be able to slap it in somewhere and then I will deal with reporting later. And so that gives those of us who have been working with SQL Server the heebie-jeebies a little bit. So again, we’re not necessary looking for, the secret sauce or any secret information here, but I guess I am looking for your thoughts. You’ve dedicated 26 years to SQL Server.
Bob: Yeah, 26 years, plus, now, almost 27.
Carlos: That’s right. So, what’s your thoughts on the evolution of SQL Server? Like who is going to be using it? And admittedly, so those of us who are listening to the podcast, supporting it, developing with it, is that going to change? Are only big financial institutions going to be using it? What’s your take there?
Bob: That’s a fair question, because of all the different form factors we live in. You first take the Pi example, we specifically built that because of developers. I mean, developers came to us and said, “this is what we have. We have IoT-type applications we’re building, but we have scenarios where we need to do processing on the device. We can’t always be able to ship raw data from these devices to say, a hub or even directly to the cloud.”
Carlos: Because there’s too much data?
Bob: Yeah, well, not just that, but sometimes they’re disconnected scenarios, or semi-connected, think about like a mining company or something. So, a great example is like a mining company. Mining company has the need to do machine learning processing and actual data processing at the mine, but not be connected all the time. So, if they have devices like this where they can have a database engine that includes things like time series and machine learning models and so forth, then we can actually put information on there. It could do processing and then connect what it needs to. And instead of sending all just the raw results, why not send processed results? So, this is developers telling us and other companies that that’s what they need, and we just said, “you know, our SQL engine can do these things. Why wouldn’t we just find a way?” So, we said, “in order to do this, we need to reduce the footprint, the memory footprint of SQL Server, the disc footprint and so forth.” So, you’re going to be blown away, but we can do this now in 300MB. SQL Server, the engine, runs in a 300MB footprint on this. So that’s crazy, given how gigabyte-sized the engine itself can take. And almost everything’s there do to the engine. So, one scenario would be is that we think developers want that kind of functionality and it’s not like all the NoSQL options just come with all of that. Think of all the richness that SQL Server just comes with this engine. You pick on SQL Server 2019 with all of the stuff that we’ve got into the product. Well, what if you need to do a DevOps scenario with containers? I just did a demonstration at Build using Azure DevOps and Pipelines where I showed how to use Azure Pipelines to automatically do SQL testing in a way for developers that have not been thought of before. And it’s a little more difficult to do that with, say, like a NoSQL system. I showed a demo where you need to go run SQL Server tests against a new cumulative update of SQL. But you don’t want to use a shared development server, because you’re always colliding with all your devs. You know these devs, they’re terrible. They’re always colliding. They’re always messing up the SQL Server, changing the configuration, slapping on a new CU and nobody needs that. So, we use Azure Pipelines, which dynamically spins up virtual machines with containers in them and runs automated tests by you simply making a small change to a file. You make a change to a file on your GitHub repository, you push that up to your GitHub repository and boom, the testing already runs automatically for you, all using containerization. We do that in SQL. In the engine team, that’s how we test SQL Server. We use Azure Pipelines and containers to do automated tests for thousands of containers out there for different SQL Server configurations, bug fixes, enhancements and so forth. So that’s just two scenarios where I think we’ve got an advantage because of the richness that the SQL Server platform provides. If you’ve heard me speak about SQL ‘19, I don’t call it an engine anymore. Certainly, the engine’s a big part of it; I call it a platform, because it includes Big Data Clusters, PolyBase, machine learning. I mean, think of all the stuff that just is there with it. So, I think we have a huge advantage to that space that it can be used for so many different things, whether it’s operational or analytical or now machine learning or Big Data. You name a NoSQL product that can really do that today. I mean, it’s not like I’m trying to compete with them, necessarily. I’m just saying that’s where the relevance I think in SQL Server has been in the past and still is, today.
Carlos: So, what I’m hearing, and Eugene and Kevin correct me if I’m wrong here, is that we still need to think about ultimately, the full pipeline or the full breadth of what it is that our applications or our pain points are, and the ability to store data is simply not the only function. Is that fair?
Eugene: Well, I mean, I think– so, I just did a course on Azure Event Hubs and the analogy that I used with that compared to a database, because I’m biased, and so my first clip is why not use a database, a normal relational database? And then one of the architects for Azure Event Hubs says maybe I took too long on the ACID compliance piece, and it’s like, “yeah, okay, that’s fair.” But the analogy I use is like comparing a luge to an off-road Jeep. Event Hubs like a luge. Like, if you’re trying to do this one thing, it is super fast, it is super performant, it is super focused, but I wouldn’t try and get groceries with it. Right? And to me, based on what Bob is saying, SQL Server feels more like an off-road Jeep, because it can handle a variety of circumstances. And even better, you may have a company where you have those different terrains, those different circumstances, and so if you can use the same tool for that Raspberry Pi, for that DevOps, for that Enterprise terabyte database, there’s some value there that’s more than just the individual solutions.
T-SQL is a major language for data pros
Bob: Absolutely. Also just think about the T-SQL language. You know, I don’t know if you noticed or not, but when we were looking at machine learning services, we were trying to study, “what should we do?” So, we’d start with R and then we did Python, then we’re like, “what should we do next?” And somebody said, “that’s probably okay, because T-SQL is actually the third most popular language for data scientists in the world.” So, the T-SQL language is across everything we do with SQL Server, whether it’s in Azure, Linux, containers, Edge, you name it. And then look what we do with it, we just extend that language. Graph database is a developer feature we offered right in SQL ’17. Kind of an underrated feature, in my opinion. It’s pretty good. We actually enhanced it in ’19, as well. It’s all T-SQL-based because people said, “I don’t want to have to go another language for that. You use PolyBase, it’s T-SQL external tables, you do machine learning, it’s all based on T-SQL. So, I think it’s underrated, in my opinion, the fact that we have a standardization for a language and interface that can do so many things like Eugene just talked about.
Carlos: And then able to divide some of those tasks So, I would think that there is a good portion of those people that can do T-SQL that don’t necessarily consider themselves developers.
Bob: Right, but they are. They probably don’t realize it, right?
Carlos: No, no, no, that’s right.
Bob: Look at the NoSQL systems that always pivot to SQL as a language, have you ever seen that? Everybody’s got a SQL adapter.
Eugene: Kevin, don’t you have a law about that? I think Kevin Feasel’s law is–
Kevin: Feasel’s Law, that every data platform technology, at it matures, eventually develops a SQL interface.
Bob: Is that not true? I mean that’s–
Eugene: Yeah, no, it is.
Carlos: Is that coined already, or can we call that the Feasel Principle?
Kevin: No, I’ve already done– I’ve had that for years.
Eugene: He’s already coined it.
Carlos: Oh, you did.
Bob: He’s copyrighted that a long time ago. I mean, it sounds kind of trite to say.
Kevin: (Unintelligible) Carlos.
Bob: It may sound trite to say that and it sounds like a marketing campaign, you know, T-SQL’s consistent, it’s true, man, think about it. Think about how consistent that language is, and we just keep extending the language, versus making you do a new language. Most people I talk to, when I really make them realize that, they’re like, “you know what? I hadn’t thought about that. That’s a pretty powerful toolbelt.” And it really makes a huge difference for developers for portability.
Carlos: No, that’s right. When you think about, I mean, we love Power BI, right, but even like Power BI, they’ve introduced new languages. And so you’re like, “oh, man.”
Bob: Yeah, those Adam Saxton people. People like that, you just can’t– what are you going to do with this Guy in a Cube person?
Eugene: You can tell them apart because they’re bald. That’s how you spot them.
Bob: Oh, man. That was Eugene, not Bob Ward.
Eugene: I know. I think they consider me a friend. I don’t know, so I think I’m friends with them.
Kevin: Probably not, anymore.
Bob: Well, I was on his podcast not long ago. I might not get invited back after that one, so.
Eugene: Yeah, I’m sorry. Guilty by association.
Carlos: Well, you can tell I’m a wannabe, right? Trying to–
Bob: Yeah, exactly.
Eugene: Well, no, I think a related thing to what we’re talking about is if you’re a database administrator, like a Microsoft SQL Server database administrator, the future looks kind of scary and it’s nice to think that okay, maybe you can ride this magic carpet into the future a little bit. You’re still going to have to start learning some Azure and some of these other pieces, but it’s nice to know that this skill set that you may have built up over decades, potentially, is not going to just be obsoleted overnight. Because I mean, you look at Azure and you look at where things are going and it’s easy to feel that way. It’s easy to feel that, “okay, I’ve gone from a Swiss Army knife that was good at a bunch of things to now, 15 different Azure scalpels that are all slightly different sizes and I have to figure out which one to use.
Bob: You know, you think about Azure, I’ve just recently been spending some time in that space and I’m currently actually working on some workshops for Azure SQL based on people that know SQL Server. Like I’m trying to help them translate that knowledge in that world, because that’s me. I’m a SQL Server person, so I started talking to my cloud team, Azure team, and started looking at it and there is so much there to learn. Trust me, as much as we pour into the cloud to help you do things, there is so much work for SQL Server DBAs, developers, administrators to still do in the cloud, it’s not a hit-a-button and you never touch it again. It’s just definitely not that. But yeah, there’s some really powerful things in it that, actually, you’ll love. Like, how many people love putting together an availability group in this complex configuration and maintaining that and getting paged in the middle of the night for it? How about avoiding some of those kind of type scenarios and focusing on other things? That’s kind of what we’re trying to do and that’s a good example of what Azure can do.
Carlos: Yeah, no question, like focus on the good stuff, and the business processes, and I always go back to that. That there are going to be circumstances that your business will need, that you’ll have to be able to understand.
Carlos: So then, as we’ve kind of gotten to 2019, and it’s hard to pick amongst all your children, but do you have a favorite feature, or do you have something that you’re like most geeked out about, maybe?
Bob: Oh, man, my developers are all listening to this. I’m gonna make somebody mad, right? So, being an engine guy from so way back, like, I don’t know if you guys remember, I used to do a lot of engine internal talks. Do you guys remember some of those, Inside Latching and those kinds of things? I still get questions about that all the time, and in fact, I’ll make a quick plug for these guys, there’s something called EightKB Conference coming up.
Eugene: Oh, yeah.
Kevin: Yep, actually, hosted in my part of the town.
Bob: Oh, is that right?
Kevin: Mark Wilkinson, out of the Triangle area, also Andrew Pruski and Anthony Nocentino.
Bob: Yeah, so Anthony was a tech reviewer of one of my books and had said, “hey, you know, we would love for you to do this event.” And I said, “let me revive my SQL Memory talk.” So, I’ve got like a memory internals talk I’m doing as part of that event, and that’s fun for me, cause I don’t do those as much, anymore, but again, I’m an engine guy, tried and true, even though I love all these great features around it. So, there’s this feature called Accelerated Database Recovery, have you all seen that one?
Bob: Okay, so let’s have a little fun here on the phone. So, Carlos kicks off this transaction and deletes a billion rows in a transaction, he just gets off and clicks this–
Carlos: Cause I don’t believe in WHERE clauses, right?
Bob: He makes a mistake, right? So, Kevin is really, really mad, so he tries to kill Carlos. I mean, kill his session. So, he goes in and he runs the Kill command to kill Carlos’s session, and it goes back into this rollback state, so he really wants to kill him now, like he’s really mad at him. And the business is down at this point, it’s really, really bad. So, Eugene comes along and says, “you know, I read on the internet that Paul Randall said that if you just crash SQL Server, it’ll just magically come back up in this scenario.” So, Eugene goes off and crashes SQL Server and it comes back up and it goes in the Error Log and it says 1%, 2%, 3% or whatever, and the whole time your business is down, and I’m the guy managing the transaction log and it’s growing out of control. Does this sound familiar to you? I mean, I was in support, as you guys know, for many years, and I’m not making this up. That was a very common call for us, and customers constantly would say, “well, how long is it going to take?” And I’m like, “as long as it takes to roll this back.” So Accelerated Database Recovery literally gets rid of that entire scenario. And anytime I present it, everybody says, “this can’t be for free,” which it’s not, but you’d be surprised the impact it may not have on you. So, for example, we have this whitepaper with all the details in it, and in there we talk about the TPCE benchmark, which is kind of a common OLTP mixture-type thing. We’ve seen as low as like 3-5% impact of turning this on and running that benchmark. Now if I told you that I could like, maybe, give you a 3% hit overall and then give you all of that, you might sign up for it. Consider the fact that in Azure, we ran this on a million databases before we shipped it in SQL ’19 and by default, it’s on in Azure. Just spin up a database, we just put it on by default for you. It is the only way that we can actually achieve hyperscale, the new hyperscale functionality for Azure to be able to do that and give you the SLAs that come with it. So, it is one of the coolest, most innovative technologies I have ever been a part of at Microsoft and I just love telling that story, though. Where Carlos did something and Kevin tried to kill Ca– I mean the session, and then Eugene, you know, the whole thing. So, Paul Randall will probably hear this and send me some nasty email, because I always use him as that example. And he says to me, “well, I don’t tell people to do that.” and I’m like, “I’m sure somewhere along, you probably did tell somebody to do that.” Just kidding, Paul’s an old friend of mine, so I use him.
Carlos: What you’re forgetting is some corruption scenario.
Bob: Some corruption scenario, exactly. It is, by far, my favorite feature, because it’s just an option you turn on with your database, we do all of this stuff. What we do is we use what’s called a persisted version store in your database. It’s kind of like the version store that you have, today, for a snapshot isolation, but it’s different, but it looks similar, but it’s used for recovery purposes. And consider failovers, now; that’s the reason we do it in Azure. So, if you’ve got an availability group and you’ve got a secondary replica you’re syncing, what are we always doing? We’re always redoing transactions there. We’re trying to roll forward all the time. So, when you fail over, we have to do the undo portion. Well, if you had this long-running transaction, a failover in an availability group scenario could take a while. Well, with this it doesn’t, it just immediately comes right up, because the undo portion is so fast during this scenario. And then on your secondary replicas, the read transactions going on are using that version store, not TempDB. So, that’s very powerful, as well, for readable secondaries. So, there’s a lot of great uses of it; it’s really, really cool technology. Anybody watching this, and for you guys, also as well, go look on the internet on a search for ‘constant time recovery’. That’s the original name of the project and you’ll see this whitepaper that we wrote. And I love this whitepaper, cause you want to geek out, man? I mean, it has got all the details of the internals of how this works, all the timings and experiments that we did. but yeah, this is probably one of my big hero features that I show off all the time.
Carlos: Very cool. So then, you mentioned some of the overhead, but is this a ‘everyone should turn it on’, kind of a recommendation?
Bob: You know, we, on purpose–
Carlos: Not everyone. I mean, there’s always an exception. We have to caveat. We’re going to throw all kinds of red flags on this.
Bob: You know, in SQL Server, we have a little bit of a philosophy on these things. We’ve kind of taken a very conservative approach over the years, and not made these things defaults. Where in Azure, we feel more empowered to turn these things on and let you turn them off in the cloud. And so that’s how we’ve done that. And so, in SQL Server ’19, this is not on by default. Based on just my experiments with it, my testing with it, reading the paper and so forth, I think everyone should take a hard look at turning this on. It doesn’t mean, obviously, today I’m recommending go to your production server and just flip the database bit, though you can do that.
Carlos: Fair, fair.
Bob: But it’s something that you should take a really hard look that maybe it’s for you, especially once you make that move to ’19. It’s independent of dbcompat, too, so it’s not like you have to go to ’19 and use the new dbcompat. You can go to ’19 and use an older dbcompat and this still works. So, it’s independent of that, that’s why it’s just got such great rich functional as the engine. But again, it’s just like the feature in ’19 for TempDB we did, which we took TempDB metadata and turned it into like Hekaton tables. We don’t make that a default, either, we let you go ahead and opt in. So, SQL Server in the box, we have so many customers out there, you control your environment, we don’t. In the cloud, we kind of control the backend, so we kind of let you do opt-in. I think everybody should take a look at it, though. It’s pretty amazing stuff.
Bob: Carlos is like, “should I use this?”
Carlos: Yeah. So, I guess where I was going is, in the new Microsoft, listening to customers and bringing things in, my question was going to be, is that where you’re getting mo– let me rephrase that.
Bob: Do you want to know where this feature came from?
Bob: It came from the cloud. So, in the cloud, we use the DevOps model where the engineers themselves have to do a lot of the support work for the backend. And these guys are developers. You know? It’s like, “I don’t want to get paged in the middle of the night.” So, we were just seeing all of these problems like I just described to you where customers had these issues, and they’re like, “surely there’s a better way,” that the engine can handle this in a better way. And then also we needed to do hyperscale. We said, “oh, this hyperscale architecture requires something unique like this.” So, it was a combination of just pain points that we lived your pain, and that also we needed to innovate on things like hyperscale. That’s where the project came from. And I remember when I first joined engineering, hearing about this back in like 2016, like, “constant time recovery, what is this thing?” And I reached out to Peter Burn, a long time friend of mine in the development team, Peter’s been there forever, longer than me. I said, “Peter, what is this thing?” and he described it and I’m like, “oh my goodness, this could be life-changing.” It took us a couple years to get it to market, but yeah, it’s kind of a game-changer for me. So.
Carlos: Very cool. So, eating your own dog food, there, a little bit.
Bob: Absolutely. You’d be surprised how much in the cloud that happens, now. It’s pretty interesting, yeah.
Carlos: That’s cool. So, I know we just had Build, lots of conferences. I mean, Ignite will happen here, but in terms of, I guess where things are headed or, I don’t know, I guess I’m not asking for– well, maybe I am asking for a scoop. Any scoop you can give us, Bob?
Bob: Well, the biggest thing we did at Ignite was public preview for this bad boy. So, let me see if I can open this up. Oh yeah, here we go, yeah, here’s my Pi device, do you see that in there?
Eugene: Now is that a Pi4 you’ve got?
Eugene: I’m so mad that it requires a fan. Why does a Raspberry Pi require a CPU fan?
Bob: I am definitely not a hardware geek, but it wasn’t hard for me to put this together, but SQL Server can ta– so SQL Server Edge, we call it Azure SQL Edge, we did that for a very specific reason. We didn’t want people to think this is a full-bore SQL Server and we wanted to kind of involve– you know, when the final thing comes out, it’ll logically be like a subscription model like Azure comes with. So, it’s kind of an Azure SQL product, but it is a SQL Server engine running in here. But that was the biggest announcement from the SQL perspective we made at Build, for developers. There’s other things that we have in the plate. Remember last year at Ignite we announced something like Azure Arc. We just kind of started that journey, we’re in the middle of that journey right now, so we’re currently having private preview for customers with it. That’s an interesting beast in itself, because this is now us trying to bring the world of platform-as-a-service down to you, to your data center, to your servers. Think about like a Managed Instance that you see today in the cloud. Imagine you getting a Managed Instance in your servers. We’ll use Kubernetes as the power to make all of that work. Kubernetes provides what we think is the best software infrastructure that we know, that’s going to give us platform-like capabilities to make all of that work. Let me give you an example. So, you see today, when you go to Azure, there’s something called like Advanced Threat Protection. You may have seen some of that. Security-type models where we run machine learning models to help point out stuff to you. Imagine we could do that in your data center as a service running in your cluster with a SQL Server. Imagine you having a SQL Server in your cluster that’s what we call Evergreen or version-less. You know, if you look at Azure today, you never patch SQL, we do all of that for you. And so, imagine if you lived in that kind of world, too, where we’re constantly giving you changes or updates and we’re just kind of keeping your SQL up to date all the time. So those are the kind of things we’re looking at. People think Azure Arc as like this arc, like a circle. That’s not the case. Arc as like an electric arc. You know, we’re trying to arc together the world of the cloud and your data center in that hybrid way. We didn’t announce anything at Build for Arc, but that’s something we’re continuing to work on. But Azure Edge was our big thing at Build that we made the announcement on.
Carlos: Right. And I think about that, and in terms of accessibility, what do organizations want to do? They want to be able to develop, they want to be able to solve some of their problems. What do they not want to do? They don’t want to be patching and, you know, all of that (unintelligible).
Bob: You know, you hear both ways on that. I mean, if somebody says, “I want full control over that. I want to make sure I get to decide,” and so forth. And then you see some customers that are tired of maintenance windows on their own and having to go figure out when to patch and what to patch and doing it all themselves and patch windows and–
Carlos: So, I felt like once the– and maybe we’re going off topic, here, but I felt like once the DOD had signed on, that that was going to kind of open the floodgates and be like, “okay, well, I mean–”
Bob: To be honest, that’s a true statement. There are customers that have contacted us about Azure and said, “well, heck, if the Jedi DOD thing is going to be a thing, then maybe I should look at you guys, right? Maybe I should take a look at it.” I’m spending a lot of time on Azure right now and I really am impressed with the technology we have put behind the scenes. You look at hyperscale as well, that’s pretty amazing technology. How would you like to be able to restore a database of massive sizes of terabytes in in 5 minutes or something? Cause we’re using snapshots and those kind of type things, and yeah, it’s good stuff, and the cloud does give us ability to do things that we can’t do on-premises. You control the deployment on-premises, we don’t. But we control it in the cloud, and we have full reign over how we do that. We’ve got to keep you running, we have to meet SLAs. But since we have that ability, we can do a lot of different things that we need to do to make sure that the software is kept, maintained and enhanced. And we can do things like whitelisting; people don’t even know that. Like we can introduce a new feature, and we can call Carlos and say, “hey, we’re going to only let you use this right now, for a week and try it out.” Well, how do I do that on-premises? I can’t do that. We have to give you a special code or something. So, Arc might give you that same functionality. If you sign up with us and say, “look, I’ll be an Arc customer,” then we would just keep you up-to-date all the time and you wouldn’t even be looking at that. That’s the thinking. Bring that platform-like capabilities down to your data center. And then also we tie it together. If you allowed us, in a semi-connected or connected way, you would tie up and you would go into the Azure portal and you see all your assets, including things going in your data center. Potentially even think about scenarios where you could do pay-as-you-go SQL Servers running in your data center, versus the licensing model used today.
Carlos: Yeah, wow.
Bob: But those are the things we’re looking at. I mean, we’re always looking at new stuff for SQL Server. I mean, there’s this really smart guy named Slava Oaks that works on our team. People don’t really know who he is, right? He’s the godfather of SQL Server, so he’s one of the smartest guys. I mean, I am Penny from the Big Bang Theory when I’m in a room with Slava. Slava’s on the board, drawing, and I’m like, “what? Five? What’s the answer?”
Carlos: Are we still talking about SQL Server?
Bob: Exactly. But he’s also one of the nicest people I’ve ever met in my entire life, and so he’s such a great guy to talk to. But he is always coming up with innovation and we continue to do that, even after shipping ’19, we’re still thinking of really cool new stuff that we want to do down the road.
Is the two year cycle a good fit for coming out with new versions?
Carlos: Sure. Yeah, so so part of that breakneck pace, so we had like– you know, there’s ’14, ’16, ’17, ’19.
Bob: Yeah, we got kind of fast, didn’t we?
Carlos: So, I’m not expecting a ’20, cause we haven’t heard 2020 or 2021, so you feel like is the two year cycle like a good fit, or is that–
Bob: You know, it’s funny you mention that. I don’t know if you guys have heard of a guy named Conor Cunningham? You guys know Conor? Conor spoke a lot previously, in events in the past. Bits is one of his favorite ones he does. And you’ve seen him at PASS with me on stage, you’ve seen us doing the Dallas Cowboys stuff and everything every year. Deep friend of mine, another one of the smartest guys I know. And so, he’s our Chief Architect to release SQL Server and he tells me that if we wanted to, with our engineering systems, we could release a new version every month if we had to. We could. We feel quality-wise we could do it, but what would be in it? Seriously. So, what we’re going to do is we’re going to release new versions of SQL Server as we feel we’ve got major capabilities that somebody wants. If you look at SQL ’16, there was so much stuff we wanted to get done that was not in a ’14 or ’12 and so we packed a lot in that release. ’17 came along very quickly, primarily because of Linux. We felt we needed to get to the market with Linux in a timely fashion, and so that’s why ’17 came out right out of the gate after ’16, even though we put a bunch of other stuff in there. ’19 really was supposed to be ’18, I’ve got to be honest with you. I mean, behind the scenes, if you go read my book, I think I said that. Like, “hey, it was really supposed to be SQL ’18.” And then this big data cluster thing we built, which is an amazing set of technology, it just wasn’t ready yet, and we wanted that to be part of the release, and so we then deferred it to ’19. So, I’ll tell you one thing we’re not: we’re not on a 5-year cadence, that’s for sure. Yeah. But we would like to do things every year or so, if we could, in a major new release of SQL Server. There’s all sorts of things the team is working on, but again, we’re only going to do it in a way that we think we have something that’s really compelling to go to market that’s differentiating from what we did in the past. Customers are still just trying to put their head around SQL ’19, to be honest with you, so maybe we’ll give you a little bit of a break right now, let you breathe in what we’ve shipped in ’19.
Carlos: Yeah, let us catch up, read the book.
Bob: That’s right. Yeah, read the book, yeah, catch up and read the book. It’s only 500 pages, no big deal.
Carlos: Yeah, that’s right.
Bob: Yeah, my wife saw the book and it’s actually shorter than the Linux book I’d done, previously, and she still opens it up and she goes like, “who’s gonna– who’s gonna read this? Like, is this– is this something they’re just going to look at it?” I’m like, “no, no, no, somebody’s going to go through it,” you know? So yeah, there’s a lot there, that’s for sure.
Carlos: There you go. And so, compañeros, if you get the book, now that’s one of your responsibilities is to let us know so that Bob can tell his wife that he had some people go through the book.
Bob: That’s right. I want to hear that people like the book, because whatever time she sees stuff like this or my presentations, she’s said, “do people like this?” And I’m like, “yeah.” She goes, “there is something wrong with all of you. There something really wrong with you guys.” And I’m like, “we know that. We get it. We own it.”
Eugene: We get paid a lot of money to be messed up, tell you what right now.
Bob: We do get paid a lot of money to be messed up, I agree. I actually agree. That was a fun book for me to write. I really enjoyed that. The SQL Linux book was great, because just to let you know, before I joined Microsoft, I was a UNIX developer. We didn’t have Linux back then. I was a UNIX developer with Oracle and C++ in my older days, and when I joined Microsoft, I couldn’t believe I was joining this little PC company. Like, “why would I do PC stuff?” And so, when Linux came around, I took the opportunity to write that book, because it was fun to get back into the Linux/UNIX environment. But ’19 was just passion for me, because I was involved in the very beginning for the ’19 planning with the team. All the different meetings we had in the early days; we call it Project Seattle. That was kind of fun to name it that, Project Eres was Big Data Clusters. So, being from the very beginning and looking at the early days and the early meetings, it was fun to write that book, because for me it was a journey from end to end. All the way to launching it, to do it, and so it was a fun labor of love.
Eugene: So, I do have a question about SQL Server on Linux. I think when it was first announced, I definitely did not appreciate the evil master plan that’s now kind of come to fruition, because it makes sense that like then it came to Docker and then Big Data Clusters and now on a Raspberry Pi. And I guess something I’ve always wondered is when you all were first like moving it to Linux, did you have this full evil master plan or did you just have the idea of like, “okay, no, we want to go to Linux”? Cause when it first came out, I was like, “are you trying to like compete with Postgres, which is free? Were you hoping to like steal some Oracle customers?” I didn’t see the picture. But now it’s like, “oh, this has opened up so many opportunities for SQL Server.”
Bob: Sure. The answer is, yes to all those questions.
Bob: We want all of those customers. It’s funny, because you know, in my organization, Azure Data, we have a Postgres/MySQL team, and they’re always, talking to me like, “hey, what are you guys doing?”
Eugene: Talking trash?
Bob: This is where I’ll really pull in people like Slava Oaks and Rohan Kumar, Travis Wright, some of these people in those early days. As soon as we went to SQL on Linux, we knew we could do containers, because containerization is just a natural evolution for a Linux customer. And Docker’s not the only player in town, but container runtimes were becoming a thing as we launched SQL on Linux. They’d been around for a while, but really starting to kick in, Kubernetes starting to kick in, Brendan Burns worked for Microsoft, he’s influencing us, what we’re doing. So, we knew we could do Linux, Kubernetes, containers right out of the gate. In fact, when we launched SQL on Linux, we also announced container support right out of the gate, so we knew that in building SQL Linux.
Bob: Edge is a little different story, but it is interesting to let you know that we have been working on this for a while. I would have to go back and double check, but it’s possible we were thinking along the lines of this type of thing (Raspberry Pi) as far back as SQL on Linux, and the idea was not necessarily to be getting this, but it was, “can we do SQL on ARM devices?” We knew that was a market. Hey, if you look at SQL on Linux, just to make I make it clear, here, it wasn’t like we said, “hm, let’s go get Oracle.”
Bob: No, no, we didn’t say that. It was actually people coming to us. Red Hat, other customers said, “you know, I really like your product. In my shop, we have a lot of Windows with SQL, we have a bunch of Linux stuff. Can you do the Linux thing? Have you all thought about that?” And if you go look at that Linux book, it was this amazing idea of using this drawbridge project that Slava took on. This is when Slava came back into our organization and was specifically hired back in by Rohan to go figure out this Linux thing. And he quickly realized, “instead of trying to rewrite the entire code, I’ll use this PAL technology to quickly get to market with it.” So, once we did that, it just opened up so many doors for us. And so, when we had the confidence that we could run SQL on Linux, we said, “well, ARM processors, what the heck? We can do that.” So, the biggest challenge was not actually the ARM piece, it was the footprint of SQL Server. Robert Dorr, one of my dear friends, that’s the Other Bob of Bob SQL, he’s the smarter Bob. Again, he’s another person that I look like Penny from the Big Bang Theory with, when I get into a room with him. And he was instrumental in working with the developers to figure a way to reduce the footprint. You ever seen the movie Pearl Harbor? Where they’re on the runway, they’re trying to figure out how to launch the planes off the aircraft carrier, these big bombers. So, they’re in there like ripping everything out of the plane. You know, they’re getting all the weight out so they actually could lift off. In one scene they take the machine guns and put black broomsticks in there to make it look like they had a gun. Now this isn’t the same comparison, but it’s a similar exercise where the team went in and said, “what is all the stuff in that SQL engine over the years that nobody really cared the footprint of that we really do care?” But you still make it a very powerful thing, and so Linux gave us the confidence we could do amazing projects like this. So, yes, to your question, there is a master evil plan as part of it. It was also still a journey, though, because we didn’t even know that we really could pull this off until we really started digging into it.
Eugene: Yeah, well, I’ll be excited to use it as just for personal projects, because SQLite doesn’t allow you to link columns. I don’t know if you knew that, but–
Bob: Oh no, I’ve seen some of those kind of type restrictions. And the only thing you’ll find is in-memory OLTP is something that we couldn’t fit into this. That’s a pretty much of a hog, right?
Eugene: You have to have memory to do in-memory OLTP.
Bob: You know, but you’d be surprised this thing has 32GB of RAM and it’s like as much RAM as my laptop, here, so these things have a lot of RAM in them. But yeah, other than that, there’s some really powerful engine features that are still in here, just the same stuff that we use there in the processor all using containers.
Carlos: Very cool. So, last questions, guys, or should we do SQL Family?
Eugene: No, I’m thoroughly impressed. Very good, pleasure to get to meet you, Bob.
Bob: Well, thank you, but I just want to say one thing and I really mean this to the bottom of my heart, it’s not me, I’m just a messenger. I mean, I have this amazing, talented, nice, respectable people all up in Seattle. They’re all virtual right now, I see them all the time virtual. They don’t how I do virtually. “How’d you been doing this for years?” They are some of the coolest, smartest people I’ve ever met and they’re the ones that are behind all of this. And that’s why in the books I’m very passionate about making sure I include all of the people behind the scenes that wrote the product and worked on the product. So, you know, when you go read these books, look for those little names and stuff where I interviewed them and say, “hey, how’d you come up with this idea?” And they’re like, “oh, we were sitting around one day in the conference room and we’d heard all this customer feedback and then before you know it, it had turned into a project.” That’s how smart these people are. So yeah, thank you for that comment, but I have such an amazing team that builds this, it’s easy to write this stuff and talk about it.
Carlos: Yeah, and the program managers have been great. So, you mentioned Travis and Jos and I’m forgetting some of the ones now, but–
Bob: Well, think of Pamela Hood and Pedro Lopes and Joe Sack and Denzil Ribeiro, Ajay Jagannathan, I mean, you just go down the list of so many talented people and we’re hiring more. I just feel honored every day to work with these folks.
Carlos: So, a guestimate, right, and hopefully this is shareable information. So, how many people would you say work on the SQL Server product?
Bob: You know, it’s hard to pin that down, exactly, because since I said the SQL Server is across all of this, we have SQL Server folks that work on parts of the SQL engine that kind of focus on more of the cloud a little bit, but parts of what they do go into the SQL engine on-premises. We have people that work on the SQL Server engine and they’re starting to work on things that are going into the cloud. So, it’s not like there’s just one team called the see– in fact, I know you introduced my title, and I’ve tried to change that now to just say Azure Data, because across Azure Data, we have a lot of people working on what gets shipped as SQL Server, shipped as SQL Edge, shipped as Azure SQL. I couldn’t even begin to tell you. I’d have to look at the org chart, to be honest with you. I work with so many people. They’re all so talented, from many different types walks of what they do, both PMs and engineers, so I couldn’t even begin to tell you the number, to be honest with you. But I wouldn’t say it’s like, nuts, but it’s– there’s a large number of people working on it.
Carlos: I bring it up only to say, again, kind of circling back to the beginning, this idea of getting a little concerned about the direction of SQL Server. You think about an organization as large as Microsoft, they’re putting a pretty good investment into this.
Bob: Well, all I know is that when the Chief Financial Officer of Microsoft comes out and talks about the results of the company and your product is usually mentioned in those things, you know, you feel pretty good about the direction you’re headed. So yeah. That’s the way I look at it.
Carlos: There you go. There is that.
Bob: There is that, yeah, exactly, yeah.
Carlos: Awesome. Okay, so let’s see. Compañeros, so this is technically Season 4, which may mean nothing to you, more just organizational, probably, for us. We changed some of the advertising and things that, we’re trying to do video, change things up a little bit.
Kevin: Start of the declining season, because usually 3 is where things will peak and now it’s all downhill from here.
Bob: The declining season.
Eugene: We jumped the shark and now we’re just trying all the gimmicks.
Carlos: Well, yeah. Well, you know, trying to keep up with SQL Server, right?
Bob: Good luck with that. I can’t even keep up with SQL Server.
Eugene: The t-shirts giveaways didn’t work and now we’re doing video and book giveaways.
Carlos: Yeah, that’s right. Okay, so we have a new set of SQL Family questions.
Bob: SQL Family questions. I think I was sent these. I’m looking at them now. Interesting, okay.
Carlos: The first time that you’ve gotten them, and so no pressure. You have to let us know how you like them. Amanda, our producer, her job is only at stake if you don’t like them. No, I’m just kidding.
Eugene: It’s awkward, because she’s family, so whenever you have to fire family it gets real. Thanksgiving’s the worst.
Carlos: Yeah. No, but they’re– anyway, we hope you like them, compañeros.
Bob: Okay, it’s all right.
SQL Family Questions
Carlos: Now, we are sticking with one, so going back, all the way back in time, you gave us a little bit of history here, but in terms of how did you get started with SQL Server?
Bob: Fun story. So, 1993, my wife and I found out we were going to have our first child. I was traveling a lot for American Airlines back in those days, a subsidiary of American Airlines, and I didn’t want to travel anymore. So, I’m at a couples baby shower and a former colleague of my wife walks up to me at the party and says, “you’re a database guy, right?” And I’m like, you know, I looked at her, I’m an Oracle expert back then. I’m like, “yeah, you could say that.” She goes, “hey, well, Microsoft is hiring database people.” And I’m like, “hahaha, who would go work for Microsoft?” I mean, I just looked at her like, “I’m a UNIX server, C++ guy, you guys have that little Windo–
Carlos: “I’m a serious guy,” like–
Bob: “You have those Windows things you do, PC things.” That was the first– she had just started to work for Microsoft, and they invited me up for an interview. I went ahead and took it. I flew up to Seattle in those day. I did an interview in downtown Bellevue, Washington, and when I got back, I was like, “I don’t know.” And then the hiring manager, Andrea Stepani, who I still have occasionally talked to, recently. She called me at home and said, “hey, look, you’re like a programmer, developer. We need people to help us with us with the future of this thing. We’re going to give you access to the source code to help us look at stuff.” I’m like, “source code? Really?” And the location was ideal for me. I live in North Richland Hills, Texas, this is Irving, Texas, not a long drive. I was doing some contracting work then, I wanted a little more solid work, so I started and thought maybe I’d be there a year or so. You know? Here I am.
Carlos: You’ve run out of fingers and toes.
Bob: Exactly, exactly. So yeah, kind of a fun, fun story to get started with SQL.
Carlos: Yeah, very cool. So, you never know, I guess, you have to even do the networking at the baby shower.
Bob: Yeah, yeah, go do networking at baby showers. That will help you get a job.
Carlos: So now, having said all of that, now what advice would you give to someone wanting to pursue a career similar to yours? Now, admittedly, you’ve kind of– we’ve all zigged and zagged, right.
Bob: Kind of a weird career right now, right? But I did start in support, though, and you know, a lot of people maybe discount that job. That’s a harder job– well, developers are going to be mad when I say this, but they’ve said to me, you know, Paul Ramos used to say to me all the time, “you’ve got a harder job than me, man.” So, support can be a very difficult job, and Microsoft’s, I think, probably hiring right now, support people. So, support’s hard, because you’re combining those customer service skills with deep technical knowledge. A lot of people don’t realize that you can get involved in a company like Microsoft doing technical support. Microsoft has something called a Mock Program where they hire people to go do those kinds of things, so that’s one consideration, if you just like technology. I would say still though, today, think about Azure and the cloud. As much of a big deal as SQL Server is, if you start getting experience, you start studying Azure and getting knowledge of that and you walk into somebody and says, “I know Azure,” you’ve got a good shot at getting hired, to be honest with you. Or Amazon, Amazon web services, you know, pick a cloud technology like that and then combine that with say like database skills or development skills and I would recommend that to anybody right now trying to launch their career or start their career.
Carlos: There you go. What’s one thing that can instantly make your day better?
Bob: Oh, that’s an easy answer. Looking at my wife, I mean, seeing my wife every morning. But it’s really true.
Kevin: Did she just walk into the room?
Bob: No. Yeah, “hey Ginger.” My beautiful wife, Ginger, is this typical sweet Southern woman, but tough as nails and perhaps the nicest person in the entire world. And waking up and seeing her smiling and she’s always got a you know, bright smile, especially even during our virus situation and all the things that are going on, she’s kept such a great attitude during all of it. Funniest thing she said to me, though, a couple weeks ago, she said, “now when do you travel again? You’re here a lot.” So, but no, absolutely Ginger. That’s an easy answer.
Carlos: There you go. Now, you have $10k. What computer or tech would you purchase?
Bob: You know, I’m just an old school SQL guy, right? Like I would love to get myself a souped-up server, I think. I want my own little data center. Can I have a data center? Can I get the Ward Data Center? I’ll start renting it out.
Carlos: Yeah, there we go. I was going to say, “I’m not sure if $10k would get you there.”
Bob: I don’t know, man, you’d be surprised these days. I think you could get something pretty massive. I don’t know. I would have to do like a s–
Kevin: Yeah, or eBay, yeah.
Carlos: Or eBay.
Bob: Or eBay, yeah.
Eugene: Get a (unintelligible).
Bob: Maybe get another laptop or something? I don’t know, I’m not a tech head as much as like– you know, Adam Saxton of course, would probably tell you he wants an entire studio recording super uber duper camera. You know, I don’t know if you know Adam, but he’s all full of monitors, too. So maybe I should get that, like a huge monitor, bigger than I have right now that I’m looking at. So maybe that’s what it is, I’ll get a new, cool, widescreen, Adam Saxton-bending monitor or something like that. How does that sound?
Carlos: So, the next time you’re on this podcast you can be like, “hey, look at what I got.”
Bob: Yeah, exactly, exactly. You can tell Adam I said that, too.
Carlos: Okay. So which famous person in history would you want to spend the day with?
Bob: There is no question in my mind, it would be Jesus. I would love to be like, with the Apostles. I just want to see all of that happen, just up close. I’m a deep Christian man, deep of faith and to be part of that history and to see that live would be absolutely ama– just to get to ask him a question. “Hey, can I ask you a question?” Some really wise question, you know He’ll come back with the wise answer, obviously. So yeah, that would be the person, for sure.
Eugene: I don’t know. It’s funny, you say about asking Him a question, but it seems like at least in the Bible, like half the time the disciples are like, “what did you say?”
Bob: Oh, I know. That’s what’s going to happen. I’m going to ask a question and He’s going to come back with something else and I’m like, “can you do a parable real quick on that one, cause I– I–?” And then I won’t get that and then it’ll take a while and then He’ll say, “oh, you don’t have enough faith, Bob.” And I’m like, “yeah, you’re right, I know, Jesus, I don’t.”
Carlos: There you go.
Bob: But yeah, it would be Jesus, for sure.
Carlos: If you haven’t read the Kingdom and the Crown, I will forward that series to you.
Bob: Okay, please do, yeah.
Carlos: It’s a very interesting take.
Bob: Awesome, yeah.
Carlos: Okay, and our new– well, maybe not last question, but our last SQL Family question is what’s your favorite ice cream topping?
Bob: Favorite ice cream topping?
Eugene: Yeah, don’t say ‘nothing’.
Bob: I want to say, like, whipped cream and nuts or something. I like sprinkled nuts with any kind of ice cream. So, I love combining peanuts with a lot of stuff like that, so it would be sprinkled nuts and whipped cream or something, probably, yeah.
Carlos: There you go. Can’t go wrong with that. Bob, thanks so much for being on the program today.
Bob: I’m honored to be here. Thanks for having me, I’m really serious about the books, Carlos, just contact me offline and I will donate the books as you guys see accordingly to give to your listeners and I’ll be happy to sign all of them, as well.
Carlos: There you go. Yeah, so we will make that available at sqldatapartners.com/bobward. We’ll have the show notes for today’s episode, we’ll have like the links for the constant time recovery and some of the other things we talked about in today’s episode, and a way for you to sign up to get the books. First come, first serve, compañeros, so if you listen right when the show comes out, hats off to you and you’ll have access to the book. Okay, so as we wrap up, we do want to give a little love to some compañeros out there. Showing a little love for the podcast, Ruben Escobar down in Costa Rica. Cono Babino loves the podcast and then he’s like, “well, I want to talk about Snowflake.” I’m like, “Cono, okay, but.”
Bob: Okay, Cono.
Carlos: You know, I have to stop drinking my Microsoft Kool-Aid for a little while, I guess. Warren Dean, again, giving us some love on the podcast and Gino Sena. Gino’s a new DBA, learning from the podcast, so thanks Gino for reaching out. Okay, so as always, compañeros, we are interested in your topics and what you want to hear. We mentioned there that Cono wants to talk about Snowflake. We have gotten lots of requests for more traditional engine-type things, so Bob would be proud of us for going back to some of that. We’ve been doing a lot of Power BI lately and so we promise that we will return to our SQL Server roots here and talk about some of that stuff.
Carlos: But we haven’t forgotten all about Power BI. We still have some Power BI episodes–
Eugene: People probably think it’s my fault. You’re the one who’s always pushing it. I just got an email from Amanda like, “we need more Power BI topics.” I’m like, “I’m running out of ideas, here.”
Carlos: Yeah, so this is the dirty little secret. Well, so Bob knows, probably. So, it’s great talking about this stuff, but coming up with all of that content?
Bob: Oh yeah, that’s the challenge, always, yes, yes.
Carlos: Oh my gosh, I need to reach out to people who have it all together. Right? Like that’s always the big piece. Anyway, so if people wanted to reach out to you, Bob, how would they do that?
Bob: [email protected], and you can follow me at Twitter. I’m not Adam Saxton, I’m not this big Twitter guy, but I do get out there a little bit. So, @bobwardms and just Bob Ward for Microsoft on LinkedIn. I forgot the exact LinkedIn profile, but I’m the only Bob Ward for Microsoft on LinkedIn.
Carlos: There you go.
Bob: I’d love to hear from you.
Eugene: Yeah, you can find me @sqlgene on Twitter and sqlgene.com.
Eugene: He took a strategic drink of water.
Bob: Perfect timing, perfect timing.
Kevin: If you open up Bob’s book on SQL Server 2019, go to the PolyBase chapter, the second letter of the fifth word of every paragraph, combined together will give you access to the IRC channel that I hang out in.
Bob: Kevin, you gave away our secret. The National Treasure code. The National Treasure code, you gave it away. Gosh, Kevin, you shouldn’t have said that.
Carlos: Now he’s not going to invite you back, Kevin.
Bob: That’s hilarious. That’s actually a pretty cool answer.
Eugene: Kevin never gives a straight answer for that question.
Bob: It’s a good one, though. I had to think about it for a second. I’m like, “did I do that?” That would be–
Kevin: Try it and see.
Bob: Yeah, give it a shot. I’m going to be after the thing going, “wow, it really works.” Kevin was, you know, subliminal when I was writing it. “Bob, put this in there.” That’s awesome.
Carlos: That is hilarious. And compañeros, thank you again for showing the podcast a little love and for giving us feedback on things you think we should be talking about. You can reach out to me on LinkedIn. I’m at Carlos L Chacon. I think that’s going to do it for today’s episode. we’d like to thank Bob again for coming on. And Eugene and Kevin for helping me out, here. And compañeros, we’ll see you on the SQL Trail.
Listen to Learn
00:27 Intro to the guest & topic
05:45 Book Giveaway
07:24 Get your very own SQL Data Partners Podcast t-shirt!
08:10 Intro to the topic
09:57 Mining companies are one reason why the footprint size has changed
12:45 The ability to store data is not the only function
14:09 T-SQL is the major language in all things SQL Server
15:21 Kevin has a law about how data platform technologies evolve
16:39 So many DBAs see a scary future for their jobs
18:25 Bob’s favorite feature of SQL Server 2019
19:29 Fun little story about why you should use Accelerated Database Recovery
21:48 Geek out on the whitepaper about Constant Time Recovery
23:04 Should everyone be using Accelerated Database Recovery?
24:46 Where the Accelerated Database Recovery feature came from
25:49 What’s coming up from Microsoft?
28:16 What customers want and what they don’t want
31:00 Is the two year cycle a good fit for coming out with new versions?
32:57 There something really wrong with you guys
34:39 The full evil master plan…
38:56 Bob has an amazing team full of incredibly smart people
40:11 Just how many people are working on the SQL Server product?
41:46 SQL Family Questions
49:25 Compañero Shout-Outs
50:30 Closing Thoughts & contact information
Just finished listening to episode 200 and loved it! I passed the link along to my whole team.
Bob was so gracious with his time. Lots of good stuff there. Glad you enjoyed it!