1 00:00:00,000 --> 00:00:02,540 Welcome to the Cloud Native in 15 Minutes podcast. 2 00:00:03,200 --> 00:00:05,820 If you haven't listened before, the goal of the podcast is simple, 3 00:00:06,240 --> 00:00:10,060 to explain some complex but increasingly important and even strategic technologies 4 00:00:10,060 --> 00:00:14,320 in about the time it takes for a coffee break or a drive to the office or a walk 5 00:00:14,320 --> 00:00:16,360 or whatever it is you do in 15 to 20 minutes. 6 00:00:16,960 --> 00:00:20,440 We're all busy, but that's not an excuse to not understand, at least at a high level, 7 00:00:20,520 --> 00:00:25,020 some of the fundamental technologies that will be powering your company for the next decade. 8 00:00:25,020 --> 00:00:29,720 So I am Derek Harris from Pivotal, and the topic today is SQL, 9 00:00:30,260 --> 00:00:35,380 specifically new slash distributed slash next-gen SQL, 10 00:00:35,540 --> 00:00:39,180 the type of databases that are focused on scale, simplicity, performance, 11 00:00:39,400 --> 00:00:41,880 and quite often geographic distribution. 12 00:00:42,540 --> 00:00:46,560 Joining me to explain this new class of SQL databases is Peter Mattis, 13 00:00:46,920 --> 00:00:51,900 the co-founder and CTO of Cockroach Labs, the creators of CockroachDB. 14 00:00:51,900 --> 00:00:56,820 And I should first issue a disclaimer where we ran into some audio issues recording 15 00:00:56,820 --> 00:00:59,180 and then attempting to re-record this episode. 16 00:00:59,400 --> 00:01:01,200 So please bear with us. 17 00:01:01,240 --> 00:01:02,300 It's not unlistenable. 18 00:01:02,420 --> 00:01:03,460 It's just not ideal. 19 00:01:04,460 --> 00:01:10,000 I posted a cleaned up transcript of this episode at cloudnative.simplecast.com 20 00:01:10,000 --> 00:01:12,960 and will include it in the accompanying blog post as well 21 00:01:12,960 --> 00:01:15,720 if you prefer to take in the content that way. 22 00:01:16,000 --> 00:01:17,240 As for the content itself, 23 00:01:17,720 --> 00:01:20,580 Peter explains why you're hearing so much right now about, you know, 24 00:01:20,580 --> 00:01:23,180 these newer next-gen or distributed SQL databases, 25 00:01:23,600 --> 00:01:25,480 a group that includes CockroachDB, 26 00:01:25,680 --> 00:01:29,020 but also offerings from other startups such as YugoBite 27 00:01:29,020 --> 00:01:31,240 and all three major cloud providers. 28 00:01:31,740 --> 00:01:36,120 He also dives into the connection between these newer SQL databases 29 00:01:36,120 --> 00:01:38,360 and the cloud native world, specifically Kubernetes, 30 00:01:39,000 --> 00:01:42,720 and why large enterprises should feel comfortable trusting open source startups 31 00:01:42,720 --> 00:01:44,120 for their mission-critical applications. 32 00:01:44,460 --> 00:01:48,740 So stick around and listen to Peter Mattis of CockroachDB now. 33 00:01:48,740 --> 00:01:51,420 Okay, so let's get into it. 34 00:01:51,600 --> 00:01:56,880 Peter, can you explain in about 30 seconds what Cockroach Labs and or CockroachDB are 35 00:01:56,880 --> 00:02:00,180 and kind of, you know, just this idea of distributed SQL in general? 36 00:02:01,060 --> 00:02:03,740 Yeah, the idea behind distributed SQL came out 37 00:02:03,740 --> 00:02:06,700 because we noticed there's like kind of a gap in the landscape. 38 00:02:07,660 --> 00:02:10,660 There's these NoSQL databases, which were coming for a company for a company, 39 00:02:10,660 --> 00:02:14,760 and they were promising easy, horizontal scalability. 40 00:02:15,260 --> 00:02:20,300 And yet they tossed out so many, you know, easy-to-use functionality that developers like, 41 00:02:20,360 --> 00:02:21,680 like indexes and transactions. 42 00:02:22,160 --> 00:02:23,480 And on the other side of the spectrum, 43 00:02:23,480 --> 00:02:27,300 there are traditional SQL databases that provide indexes and transactions 44 00:02:27,300 --> 00:02:28,960 and all sorts of other goodness, 45 00:02:28,960 --> 00:02:31,020 but they didn't have that horizontal scalability. 46 00:02:31,440 --> 00:02:34,920 So we wanted to build a system that had the best of both worlds. 47 00:02:35,240 --> 00:02:41,120 Easy, horizontal scalability with all the SQL goodness that makes developer lives nice. 48 00:02:41,580 --> 00:02:46,060 And what's happening today, I guess, like business-wise, technology-wise, et cetera, 49 00:02:46,180 --> 00:02:49,080 that makes, because there's a bunch of, you know, 50 00:02:49,180 --> 00:02:52,220 let's call them new SQL or, you know, variants on SQL, right? 51 00:02:52,220 --> 00:02:56,660 What's happening right now that makes these new variants on SQL databases so valuable? 52 00:02:57,300 --> 00:03:02,260 Well, the thing that's, I would look to not so much as valuable as possible. 53 00:03:02,360 --> 00:03:04,680 Why are these all coming out kind of near the same period of time? 54 00:03:05,200 --> 00:03:09,380 It's like, why didn't, you know, multiple people independently come up with the idea of the atomic bomb 55 00:03:09,380 --> 00:03:12,140 and groups come up with, they kind of come with these breakthroughs? 56 00:03:12,400 --> 00:03:16,740 Well, there's a kind of technological change that occurred over the past 40 years 57 00:03:16,740 --> 00:03:18,240 since databases have been in existence. 58 00:03:18,460 --> 00:03:20,580 And that is, you know, computers have gotten faster. 59 00:03:20,680 --> 00:03:26,060 We all know about Moore's Law, but networks have even outpaced the improvements in processors. 60 00:03:26,460 --> 00:03:30,260 And now it's, you know, I can read data from a remote machine, 61 00:03:30,440 --> 00:03:34,520 especially one that's nearby in a data center, faster than I can read it from disk. 62 00:03:34,980 --> 00:03:35,960 That's pretty incredible. 63 00:03:36,660 --> 00:03:38,960 Not a lot of people realize it outside the industry. 64 00:03:39,560 --> 00:03:40,900 Inside the industry, it's well known. 65 00:03:41,700 --> 00:03:44,600 And then the other big change is that we actually have a much better understanding 66 00:03:44,600 --> 00:03:49,480 of how to do, you know, distributed transactions and distributed consensus. 67 00:03:50,080 --> 00:03:54,480 All this stuff has come out in the past 20 years about Paxos and Raft, 68 00:03:54,560 --> 00:03:56,160 which are distributed consensus protocols. 69 00:03:56,380 --> 00:03:59,620 And then there's distributed transaction protocols that exist on top of that. 70 00:03:59,860 --> 00:04:02,720 So I think there's like a realization by a whole bunch of different people 71 00:04:02,720 --> 00:04:05,820 that you can put these things together and make a distributed SQL system. 72 00:04:06,320 --> 00:04:08,900 And what we're seeing is, you know, various different, you know, 73 00:04:08,940 --> 00:04:12,300 kind of ways to piece together these components into a working system. 74 00:04:13,420 --> 00:04:13,900 All right. 75 00:04:14,140 --> 00:04:16,960 Are there specific application types that benefit from that? 76 00:04:17,040 --> 00:04:20,920 I mean, I'm just curious, like, why I'm interested in using this. 77 00:04:21,280 --> 00:04:21,460 Yeah. 78 00:04:21,600 --> 00:04:25,980 So if you were trying to use like a NoSQL system, like any of the applications 79 00:04:25,980 --> 00:04:29,400 that were being built on top of NoSQL systems, you know, 80 00:04:29,420 --> 00:04:32,120 the systems I'm thinking here, like inside Google would have been Bigtable 81 00:04:32,120 --> 00:04:35,760 and then outside, say, you know, systems like Cassandra and HBase. 82 00:04:36,360 --> 00:04:39,240 You can build those on top of something like CockroachDB as well. 83 00:04:39,860 --> 00:04:42,900 And the workloads those work well for are ones that, you know, 84 00:04:42,920 --> 00:04:44,380 you're doing a mixture of reads and writes. 85 00:04:44,880 --> 00:04:47,860 CockroachDB provides better transaction semantics for those workloads. 86 00:04:48,520 --> 00:04:50,980 So some things that might have been awkward to do in Cassandra 87 00:04:50,980 --> 00:04:53,260 are a little bit more natural to do in CockroachDB 88 00:04:53,260 --> 00:04:55,460 or significantly more natural to do in CockroachDB. 89 00:04:57,140 --> 00:04:59,360 Then one of the things that a distributed SQL, 90 00:04:59,580 --> 00:05:03,300 in particular CockroachDB's distributed SQL unlocks, 91 00:05:03,300 --> 00:05:06,120 is the ability to do multi-region clusters 92 00:05:06,120 --> 00:05:08,560 and have geo-distributed replication. 93 00:05:09,220 --> 00:05:13,640 With the emergence of applications that have global concerns, 94 00:05:13,940 --> 00:05:16,000 global concerns in the sense of, you know, 95 00:05:16,080 --> 00:05:18,540 regulatory compliance concerns like GDPR 96 00:05:18,540 --> 00:05:20,900 or just users that are spread throughout the world, 97 00:05:21,300 --> 00:05:23,060 people who have a database that, you know, 98 00:05:23,100 --> 00:05:25,640 scales and can adapt to where those users are 99 00:05:25,640 --> 00:05:28,500 is an important bit of new functionality that we provide. 100 00:05:29,280 --> 00:05:29,760 Okay. 101 00:05:30,200 --> 00:05:31,940 And I mean, you touched on this, 102 00:05:31,980 --> 00:05:33,220 but I wanted to ask specifically, 103 00:05:33,760 --> 00:05:36,040 is something like CockroachDB or, you know, 104 00:05:36,140 --> 00:05:38,040 pick something similar on the market, 105 00:05:38,160 --> 00:05:40,920 like a replacement for an existing Postgres 106 00:05:41,660 --> 00:05:44,560 or MySQL plus NoSQL databases, 107 00:05:44,840 --> 00:05:46,580 or are they targeting really, 108 00:05:46,780 --> 00:05:48,380 like you mentioned, kind of different workloads? 109 00:05:50,640 --> 00:05:51,880 Oh, they're absolutely a replacement. 110 00:05:52,380 --> 00:05:55,740 I mean, if you were to use Postgres or MySQL directly, 111 00:05:55,860 --> 00:05:57,160 I mean, these are functional, you know, 112 00:05:57,740 --> 00:05:59,180 kind of battle-tested databases, 113 00:05:59,620 --> 00:06:02,180 and yet setting up the replication is a headache. 114 00:06:03,260 --> 00:06:04,800 Dealing with the failover is a headache. 115 00:06:05,000 --> 00:06:06,260 And dealing with that, the scalability, 116 00:06:06,480 --> 00:06:08,840 you know, once you get to a certain size limit 117 00:06:08,840 --> 00:06:10,780 for Postgres or MySQL instance, 118 00:06:11,120 --> 00:06:12,980 you know, you actually end up having to do sharding. 119 00:06:13,300 --> 00:06:15,300 And that sharding is this huge burden 120 00:06:15,300 --> 00:06:16,440 on application developers. 121 00:06:17,020 --> 00:06:19,160 It requires them to put logic in their application 122 00:06:19,160 --> 00:06:20,800 to worry about how I'm spreading my data 123 00:06:20,800 --> 00:06:22,160 across multiple, you know, 124 00:06:22,200 --> 00:06:23,480 Postgres or MySQL instances. 125 00:06:24,280 --> 00:06:26,300 So any of those workloads that you were putting on, 126 00:06:26,380 --> 00:06:27,800 you know, Postgres or MySQL before, 127 00:06:27,900 --> 00:06:28,880 you can put on CockroachDB, 128 00:06:29,140 --> 00:06:30,560 and we just handle that sharding for you. 129 00:06:30,640 --> 00:06:32,620 We handle all the replication just out of the box. 130 00:06:33,360 --> 00:06:36,800 The availability story is just much, much smoother. 131 00:06:37,580 --> 00:06:37,980 Okay. 132 00:06:37,980 --> 00:06:39,880 And is this a matter of, 133 00:06:40,540 --> 00:06:43,360 I mean, do the users typically come 134 00:06:43,360 --> 00:06:44,700 at a certain size or scale? 135 00:06:44,820 --> 00:06:45,580 I would imagine, right, 136 00:06:45,620 --> 00:06:47,000 if I'm a two-person shop, 137 00:06:47,340 --> 00:06:49,080 I might or might not need, you know, 138 00:06:49,100 --> 00:06:50,600 that level of replication, distribution, 139 00:06:50,780 --> 00:06:52,000 or is, I mean, just, yeah. 140 00:06:52,100 --> 00:06:54,600 So where is, like, the usual user, right? 141 00:06:54,600 --> 00:06:55,620 What's that profile look like? 142 00:06:56,180 --> 00:06:58,060 You know, we see a kind of mix. 143 00:06:58,240 --> 00:06:59,540 You know, we see people who have 144 00:06:59,540 --> 00:07:00,740 a relatively small amount of data 145 00:07:00,740 --> 00:07:02,120 and just want to make sure that, you know, 146 00:07:02,140 --> 00:07:04,060 that, you know, the high availability story 147 00:07:04,060 --> 00:07:06,140 that you provide is much, much cleaner. 148 00:07:06,600 --> 00:07:08,800 And we also see people in shops 149 00:07:08,800 --> 00:07:10,140 that are, you know, talking about, 150 00:07:10,200 --> 00:07:11,620 you know, very huge workloads. 151 00:07:12,820 --> 00:07:14,540 I mean, like, if you're just a two-person shop, 152 00:07:14,740 --> 00:07:15,660 one of the things that, you know, 153 00:07:15,900 --> 00:07:17,700 the industry as a whole is moving towards 154 00:07:17,700 --> 00:07:19,320 is database as a service. 155 00:07:19,440 --> 00:07:20,920 You're providing the database as a service. 156 00:07:21,480 --> 00:07:22,920 You know, if you're a two-person shop, 157 00:07:22,980 --> 00:07:24,020 you don't want to be running Postgres 158 00:07:24,020 --> 00:07:25,020 or MySQL yourself. 159 00:07:25,380 --> 00:07:26,800 You're probably wanting to run, you know, 160 00:07:27,140 --> 00:07:29,880 RDS or whatever the equivalent is on Google. 161 00:07:30,660 --> 00:07:32,020 And we're also offering 162 00:07:32,020 --> 00:07:33,600 Cockroach as a service 163 00:07:33,600 --> 00:07:36,920 via a, you know, managed service offering. 164 00:07:37,100 --> 00:07:37,920 That's currently in beta 165 00:07:37,920 --> 00:07:40,620 and we expect to launch it 166 00:07:40,620 --> 00:07:42,880 into wider availability this fall. 167 00:07:43,580 --> 00:07:44,120 All right. 168 00:07:44,580 --> 00:07:46,500 And if I decide to adopt 169 00:07:46,500 --> 00:07:47,740 a distributed SQL database, 170 00:07:47,840 --> 00:07:49,800 what do I need to do differently 171 00:07:49,800 --> 00:07:50,960 in terms of managing it, 172 00:07:51,000 --> 00:07:51,860 in terms of writing to it? 173 00:07:51,920 --> 00:07:54,600 I mean, I assume it's not a one-to-one mapping 174 00:07:54,600 --> 00:07:55,560 in terms of skills, 175 00:07:55,740 --> 00:07:56,580 but, you know, 176 00:07:56,620 --> 00:07:57,680 maybe you're going to tell me differently. 177 00:07:59,760 --> 00:08:00,120 Yeah. 178 00:08:00,280 --> 00:08:01,900 Well, there is a lot of overlap 179 00:08:01,900 --> 00:08:03,180 and that's one of the reasons 180 00:08:03,180 --> 00:08:04,740 that we actually adopted a SQL 181 00:08:04,740 --> 00:08:07,080 as the interface language for CockroachDB 182 00:08:07,080 --> 00:08:08,700 is single as lingua franca 183 00:08:08,700 --> 00:08:11,040 for data manipulation industry. 184 00:08:11,040 --> 00:08:13,100 Now, there are just hundreds of thousands, 185 00:08:13,300 --> 00:08:14,700 millions of developers who know SQL 186 00:08:14,700 --> 00:08:16,760 and know the intricacies of SQL. 187 00:08:16,900 --> 00:08:18,560 And most of those just map perfectly 188 00:08:18,560 --> 00:08:19,880 onto CockroachDB. 189 00:08:20,560 --> 00:08:22,100 The thing you do have to be aware of 190 00:08:22,100 --> 00:08:22,720 is when you're going to need 191 00:08:22,720 --> 00:08:23,920 to multi-region deployments 192 00:08:23,920 --> 00:08:26,640 that you are having to be concerned 193 00:08:26,640 --> 00:08:28,740 about the inner-region latencies. 194 00:08:29,240 --> 00:08:31,280 We don't violate the laws of physics. 195 00:08:31,820 --> 00:08:32,140 They're, you know, 196 00:08:32,240 --> 00:08:34,160 your data centers are, you know, 197 00:08:34,180 --> 00:08:35,260 on different sides of the continent. 198 00:08:35,400 --> 00:08:36,820 There's going to be latencies between them. 199 00:08:38,060 --> 00:08:39,620 But when you know about those latencies, 200 00:08:39,620 --> 00:08:40,820 you can take them into account 201 00:08:40,820 --> 00:08:41,600 in your schema design. 202 00:08:41,920 --> 00:08:43,600 You know, it's essentially equivalent 203 00:08:43,600 --> 00:08:45,760 to the realization that, 204 00:08:45,980 --> 00:08:47,400 oh, I need to know about the concept 205 00:08:47,400 --> 00:08:49,160 of index inside a database 206 00:08:49,160 --> 00:08:50,600 and what impact that's going to have 207 00:08:50,600 --> 00:08:51,260 on performance. 208 00:08:51,580 --> 00:08:53,160 There's a similar bit of concern 209 00:08:53,160 --> 00:08:54,220 with regards to, like, 210 00:08:54,280 --> 00:08:55,900 the geo-distribution of your data. 211 00:08:56,180 --> 00:08:57,300 Need to take into consideration, 212 00:08:57,640 --> 00:08:58,320 but it's something people 213 00:08:58,320 --> 00:08:59,240 can wrap their heads around. 214 00:09:01,820 --> 00:09:02,380 All right. 215 00:09:02,880 --> 00:09:05,120 And why not, you know, 216 00:09:05,160 --> 00:09:05,780 just use, I mean, 217 00:09:06,620 --> 00:09:08,120 like there are so many cloud providers 218 00:09:08,120 --> 00:09:09,020 out there offering, you know, 219 00:09:09,040 --> 00:09:11,260 other sorts of these distributed databases, 220 00:09:11,340 --> 00:09:11,440 right? 221 00:09:11,440 --> 00:09:13,020 Why not just use, you know, 222 00:09:13,060 --> 00:09:14,760 Amazon Aurora or Cosmos 223 00:09:14,760 --> 00:09:16,020 or Google Spanner or something? 224 00:09:16,140 --> 00:09:18,700 What's the benefit of actually, 225 00:09:18,920 --> 00:09:21,000 you know, deploying a, 226 00:09:21,180 --> 00:09:22,600 whatever you want to call it, 227 00:09:22,880 --> 00:09:24,260 you know, deploying software, 228 00:09:24,380 --> 00:09:24,800 I guess, right? 229 00:09:24,800 --> 00:09:27,920 Or deploying a specific database. 230 00:09:28,580 --> 00:09:28,820 Yeah. 231 00:09:29,020 --> 00:09:29,620 Well, I mean, 232 00:09:29,640 --> 00:09:30,500 there's multiple, 233 00:09:30,680 --> 00:09:31,040 you know, 234 00:09:31,120 --> 00:09:32,360 kind of answers to this. 235 00:09:32,580 --> 00:09:32,740 I mean, 236 00:09:32,740 --> 00:09:34,160 one is we're offering capabilities there 237 00:09:34,160 --> 00:09:35,340 that you don't get from Spanner 238 00:09:35,340 --> 00:09:36,260 or from Aurora 239 00:09:36,260 --> 00:09:38,440 yet in our geo-distribution capabilities. 240 00:09:39,480 --> 00:09:40,600 The other big answer 241 00:09:40,600 --> 00:09:42,880 that affects a ton of customers 242 00:09:42,880 --> 00:09:44,800 is that there's that vendor lock-in. 243 00:09:45,000 --> 00:09:45,900 If you're going with Aurora 244 00:09:45,900 --> 00:09:47,220 or you're going with Spanner, 245 00:09:47,340 --> 00:09:48,480 you're locked into those vendors. 246 00:09:48,940 --> 00:09:49,800 And that can be 247 00:09:49,800 --> 00:09:50,800 a significant hardship, 248 00:09:51,180 --> 00:09:52,060 especially if you're trying 249 00:09:52,060 --> 00:09:52,580 to get into, 250 00:09:52,760 --> 00:09:53,040 you know, 251 00:09:53,180 --> 00:09:54,780 these big cloud providers, 252 00:09:54,920 --> 00:09:55,800 they're providing data centers 253 00:09:55,800 --> 00:09:56,440 all over the place, 254 00:09:56,500 --> 00:09:58,100 but sometimes they have a gap 255 00:09:58,100 --> 00:09:59,520 in their coverage 256 00:09:59,520 --> 00:10:01,460 and that gap can affect your users. 257 00:10:02,640 --> 00:10:04,040 But I think it's also 258 00:10:04,040 --> 00:10:05,240 that kind of cloud lock-in 259 00:10:05,240 --> 00:10:06,320 that scares a lot of people. 260 00:10:06,780 --> 00:10:07,040 And, you know, 261 00:10:07,040 --> 00:10:07,480 this is becoming 262 00:10:07,480 --> 00:10:08,560 more and more of a concern. 263 00:10:08,960 --> 00:10:09,820 Amazon is just, 264 00:10:09,880 --> 00:10:10,060 you know, 265 00:10:10,120 --> 00:10:11,780 the huge elephant in this room. 266 00:10:11,900 --> 00:10:12,720 It's gobbling up 267 00:10:12,720 --> 00:10:13,780 kind of mindshare. 268 00:10:14,320 --> 00:10:14,560 You know, 269 00:10:14,640 --> 00:10:17,480 are they benevolent or not? 270 00:10:17,620 --> 00:10:18,280 And, you know, 271 00:10:18,480 --> 00:10:19,880 it's maybe a little bit 272 00:10:19,880 --> 00:10:20,600 too early to tell, 273 00:10:20,760 --> 00:10:22,820 but even if they are benevolent, 274 00:10:22,900 --> 00:10:24,080 they can just accidentally turn 275 00:10:24,080 --> 00:10:24,600 and, you know, 276 00:10:24,840 --> 00:10:25,420 stumble over, 277 00:10:25,780 --> 00:10:26,440 you know, 278 00:10:26,460 --> 00:10:27,000 your company 279 00:10:27,000 --> 00:10:28,100 and crush it. 280 00:10:28,100 --> 00:10:29,080 And, you know, 281 00:10:29,080 --> 00:10:29,860 I think a lot of people 282 00:10:29,860 --> 00:10:30,540 are starting to become 283 00:10:30,540 --> 00:10:31,380 concerned about that. 284 00:10:32,380 --> 00:10:32,960 All right. 285 00:10:33,320 --> 00:10:33,880 It's funny. 286 00:10:33,980 --> 00:10:34,980 So it seems like 287 00:10:34,980 --> 00:10:35,960 there's almost a, 288 00:10:36,500 --> 00:10:37,500 especially in the database world 289 00:10:37,500 --> 00:10:38,020 to some degree 290 00:10:38,020 --> 00:10:39,480 and the distributed 291 00:10:39,480 --> 00:10:40,480 and kind of next-gen 292 00:10:40,480 --> 00:10:41,600 database world, 293 00:10:41,660 --> 00:10:41,800 right? 294 00:10:41,860 --> 00:10:42,500 There are these big, 295 00:10:42,540 --> 00:10:43,440 big cloud providers 296 00:10:43,440 --> 00:10:44,660 and then there are 297 00:10:44,660 --> 00:10:45,360 a lot of startups, 298 00:10:45,520 --> 00:10:45,640 right? 299 00:10:45,680 --> 00:10:46,400 So why would, 300 00:10:47,000 --> 00:10:47,340 I mean, 301 00:10:47,440 --> 00:10:48,100 locking aside, 302 00:10:48,200 --> 00:10:49,780 why would I trust a startup 303 00:10:49,780 --> 00:10:50,980 if I'm a CIO 304 00:10:50,980 --> 00:10:51,920 or, you know, 305 00:10:51,920 --> 00:10:52,960 someone with a decision-making, 306 00:10:53,060 --> 00:10:53,700 you know, 307 00:10:53,700 --> 00:10:54,080 and budget? 308 00:10:54,180 --> 00:10:54,260 Like, 309 00:10:54,300 --> 00:10:54,900 why would I decide, 310 00:10:54,900 --> 00:10:55,300 why would I, 311 00:10:55,380 --> 00:10:56,380 why would I trust a startup 312 00:10:56,380 --> 00:10:57,400 for to do something 313 00:10:57,400 --> 00:10:58,140 mission-critical 314 00:10:58,140 --> 00:10:58,580 instead of, 315 00:10:58,720 --> 00:10:58,820 say, 316 00:10:59,180 --> 00:10:59,400 you know, 317 00:10:59,400 --> 00:11:00,060 one of these big cloud 318 00:11:00,060 --> 00:11:00,460 providers? 319 00:11:00,920 --> 00:11:01,080 Yeah, 320 00:11:01,400 --> 00:11:01,600 no, 321 00:11:01,660 --> 00:11:02,500 it's an excellent question. 322 00:11:02,760 --> 00:11:02,900 I mean, 323 00:11:02,900 --> 00:11:03,360 one of the reasons 324 00:11:03,360 --> 00:11:04,440 we were providing 325 00:11:04,440 --> 00:11:04,980 the source code 326 00:11:04,980 --> 00:11:06,640 and we went open source, 327 00:11:06,740 --> 00:11:07,560 now we have a slightly, 328 00:11:07,740 --> 00:11:08,000 you know, 329 00:11:08,000 --> 00:11:08,660 kind of modified 330 00:11:08,660 --> 00:11:09,660 where our license 331 00:11:09,660 --> 00:11:10,680 isn't pure open source, 332 00:11:10,800 --> 00:11:11,540 but it still provides 333 00:11:11,540 --> 00:11:12,420 a lot of the same benefits 334 00:11:12,420 --> 00:11:13,360 is that, 335 00:11:13,420 --> 00:11:13,540 you know, 336 00:11:13,540 --> 00:11:14,220 the concern is, 337 00:11:14,480 --> 00:11:14,600 oh, 338 00:11:14,600 --> 00:11:15,200 if we just suddenly 339 00:11:15,200 --> 00:11:15,840 fold up shop, 340 00:11:15,900 --> 00:11:16,720 what happens to us? 341 00:11:17,120 --> 00:11:17,980 And this at least 342 00:11:17,980 --> 00:11:18,560 gives some, 343 00:11:18,560 --> 00:11:18,800 you know, 344 00:11:18,900 --> 00:11:19,440 modicum of, 345 00:11:19,500 --> 00:11:19,620 you know, 346 00:11:19,660 --> 00:11:20,520 people still have access 347 00:11:20,520 --> 00:11:21,560 to the source 348 00:11:21,560 --> 00:11:22,280 and a community 349 00:11:22,280 --> 00:11:23,220 can form around it 350 00:11:23,220 --> 00:11:23,820 to support it. 351 00:11:24,080 --> 00:11:24,220 I mean, 352 00:11:24,220 --> 00:11:24,740 that's kind of like 353 00:11:24,740 --> 00:11:26,620 our kind of fail-safe 354 00:11:26,620 --> 00:11:27,180 of, 355 00:11:27,240 --> 00:11:27,400 you know, 356 00:11:27,420 --> 00:11:27,960 concern. 357 00:11:29,760 --> 00:11:30,640 The other thing is, 358 00:11:30,720 --> 00:11:30,840 you know, 359 00:11:30,880 --> 00:11:31,580 like Postgres, 360 00:11:31,700 --> 00:11:32,060 MySQL, 361 00:11:32,260 --> 00:11:33,180 they're heavily invested 362 00:11:33,180 --> 00:11:35,320 by these big cloud providers 363 00:11:35,320 --> 00:11:36,460 and yet, 364 00:11:36,540 --> 00:11:36,700 you know, 365 00:11:36,700 --> 00:11:37,460 they're still open source 366 00:11:37,460 --> 00:11:37,860 themselves. 367 00:11:38,080 --> 00:11:38,220 They're, 368 00:11:38,360 --> 00:11:38,580 you know, 369 00:11:39,140 --> 00:11:40,360 to some degree, 370 00:11:40,440 --> 00:11:40,960 it's not like these 371 00:11:40,960 --> 00:11:42,180 are like the flagship 372 00:11:42,180 --> 00:11:43,880 database projects 373 00:11:43,880 --> 00:11:44,620 that people are 374 00:11:44,620 --> 00:11:46,800 putting their data on. 375 00:11:47,120 --> 00:11:47,980 They're just developed 376 00:11:47,980 --> 00:11:48,640 in the same way 377 00:11:48,640 --> 00:11:49,780 that the cockroaches. 378 00:11:50,940 --> 00:11:51,180 You know, 379 00:11:51,300 --> 00:11:52,000 we test the, 380 00:11:52,140 --> 00:11:52,680 you know, 381 00:11:53,320 --> 00:11:54,940 put significant effort 382 00:11:54,940 --> 00:11:55,560 into testing 383 00:11:55,560 --> 00:11:56,360 for correctness 384 00:11:56,360 --> 00:11:56,960 and stability 385 00:11:56,960 --> 00:11:57,860 and performance 386 00:11:57,860 --> 00:11:58,440 for our system. 387 00:11:59,320 --> 00:11:59,980 It's essentially 388 00:11:59,980 --> 00:12:00,800 like all that 389 00:12:00,800 --> 00:12:02,100 same development practices 390 00:12:02,100 --> 00:12:02,900 that they're doing, 391 00:12:03,240 --> 00:12:03,940 we do as well. 392 00:12:04,200 --> 00:12:04,380 All right. 393 00:12:04,400 --> 00:12:04,720 And then, 394 00:12:04,920 --> 00:12:05,140 I mean, 395 00:12:05,160 --> 00:12:06,280 speaking of the cloud, 396 00:12:06,380 --> 00:12:06,680 I guess, 397 00:12:06,880 --> 00:12:07,440 how would, 398 00:12:07,820 --> 00:12:08,900 how do some of these 399 00:12:08,900 --> 00:12:09,760 new SQL databases 400 00:12:09,760 --> 00:12:11,400 relate to other 401 00:12:11,400 --> 00:12:13,120 cloud native technologies 402 00:12:13,120 --> 00:12:13,680 and practices, 403 00:12:13,780 --> 00:12:13,860 right? 404 00:12:13,860 --> 00:12:14,300 You might look at 405 00:12:14,300 --> 00:12:14,940 microservices 406 00:12:14,940 --> 00:12:15,460 or Kubernetes 407 00:12:15,460 --> 00:12:16,220 or Kafka, 408 00:12:16,700 --> 00:12:17,100 serverless, 409 00:12:17,120 --> 00:12:17,720 and functions. 410 00:12:17,980 --> 00:12:18,080 Like, 411 00:12:18,440 --> 00:12:19,220 is there, 412 00:12:19,220 --> 00:12:20,000 is there a kind of 413 00:12:20,000 --> 00:12:22,200 an inherent connection 414 00:12:22,200 --> 00:12:22,880 or integration 415 00:12:22,880 --> 00:12:24,300 with some of these 416 00:12:24,300 --> 00:12:25,240 other new practices? 417 00:12:26,520 --> 00:12:26,680 Yeah, 418 00:12:26,920 --> 00:12:27,740 certainly Kubernetes 419 00:12:27,740 --> 00:12:28,420 is the one, 420 00:12:28,620 --> 00:12:28,760 you know, 421 00:12:28,760 --> 00:12:29,780 there's this huge wave 422 00:12:29,780 --> 00:12:30,860 occurring in the cloud, 423 00:12:30,940 --> 00:12:31,640 which is everybody's 424 00:12:31,640 --> 00:12:32,260 moving to Kubernetes 425 00:12:32,260 --> 00:12:33,300 and putting their 426 00:12:33,300 --> 00:12:33,900 microservices 427 00:12:33,900 --> 00:12:34,780 on top of Kubernetes. 428 00:12:35,300 --> 00:12:35,980 And for the most part, 429 00:12:36,360 --> 00:12:37,220 that works well. 430 00:12:37,760 --> 00:12:38,660 Kubernetes provides 431 00:12:38,660 --> 00:12:39,800 a great substrate 432 00:12:39,800 --> 00:12:40,700 for running 433 00:12:40,700 --> 00:12:42,300 stateless applications. 434 00:12:42,760 --> 00:12:43,540 But then we have 435 00:12:43,540 --> 00:12:44,940 the stateful applications, 436 00:12:45,160 --> 00:12:45,740 and these are the ones 437 00:12:45,740 --> 00:12:46,920 storing data in databases. 438 00:12:47,320 --> 00:12:48,220 And then you have to wonder, 439 00:12:48,320 --> 00:12:49,060 where is the database 440 00:12:49,060 --> 00:12:49,840 itself running? 441 00:12:50,540 --> 00:12:51,620 Cockroach actually runs 442 00:12:51,620 --> 00:12:52,380 really nicely 443 00:12:52,380 --> 00:12:53,540 on top of Kubernetes. 444 00:12:54,300 --> 00:12:55,840 And that is our 445 00:12:55,840 --> 00:12:56,900 number one distribution 446 00:12:56,900 --> 00:12:57,420 mechanism. 447 00:12:57,580 --> 00:12:58,360 The number one mechanism 448 00:12:58,360 --> 00:12:59,520 by which people run 449 00:12:59,520 --> 00:13:00,140 CockroachDB 450 00:13:00,140 --> 00:13:01,400 is on top of Kubernetes. 451 00:13:01,700 --> 00:13:02,300 And then they'll have, 452 00:13:02,360 --> 00:13:02,580 you know, 453 00:13:02,880 --> 00:13:03,980 Cockroach running Kubernetes, 454 00:13:04,180 --> 00:13:04,960 their application running 455 00:13:04,960 --> 00:13:05,480 on Kubernetes, 456 00:13:05,680 --> 00:13:07,000 you get all the goodness 457 00:13:07,000 --> 00:13:07,600 from the Kubernetes 458 00:13:07,600 --> 00:13:08,100 scheduling. 459 00:13:08,800 --> 00:13:09,440 One of the things 460 00:13:09,440 --> 00:13:10,420 that these kind of, 461 00:13:10,660 --> 00:13:11,840 this new wave 462 00:13:11,840 --> 00:13:12,800 of distributed, 463 00:13:12,980 --> 00:13:13,220 you know, 464 00:13:13,240 --> 00:13:14,420 SQL databases is doing 465 00:13:14,420 --> 00:13:15,300 is because we have 466 00:13:15,300 --> 00:13:15,960 the replication 467 00:13:15,960 --> 00:13:16,440 and whatnot 468 00:13:16,440 --> 00:13:17,500 just kind of built in 469 00:13:17,500 --> 00:13:18,700 and baked into the product 470 00:13:18,700 --> 00:13:20,380 that it handles 471 00:13:20,380 --> 00:13:21,580 that elastic scale-up 472 00:13:21,580 --> 00:13:22,620 that Kubernetes offers 473 00:13:22,620 --> 00:13:23,200 and it handles 474 00:13:23,200 --> 00:13:23,760 the fail-up 475 00:13:23,760 --> 00:13:24,760 that Kubernetes offers 476 00:13:24,760 --> 00:13:25,600 where you don't get 477 00:13:25,600 --> 00:13:26,340 the same behavior 478 00:13:26,340 --> 00:13:27,180 from something like 479 00:13:27,180 --> 00:13:27,600 trying to run 480 00:13:27,600 --> 00:13:28,700 MySQL or Postgres 481 00:13:28,700 --> 00:13:30,120 on top of Kubernetes. 482 00:13:31,060 --> 00:13:31,500 Okay. 483 00:13:31,760 --> 00:13:32,680 And do you see 484 00:13:32,680 --> 00:13:34,000 a common 485 00:13:34,000 --> 00:13:34,860 or at least 486 00:13:34,860 --> 00:13:36,680 a relatively common 487 00:13:36,680 --> 00:13:37,940 kind of data architecture 488 00:13:37,940 --> 00:13:38,520 as you're looking 489 00:13:38,520 --> 00:13:39,060 at customers 490 00:13:39,060 --> 00:13:39,700 and what they're running? 491 00:13:39,820 --> 00:13:41,280 The idea of Kubernetes 492 00:13:41,280 --> 00:13:42,820 as a substrate 493 00:13:42,820 --> 00:13:43,760 for running stuff 494 00:13:43,760 --> 00:13:45,140 is intriguing, right? 495 00:13:45,180 --> 00:13:45,580 So I'm curious 496 00:13:45,580 --> 00:13:46,160 if there's kind of 497 00:13:46,160 --> 00:13:48,260 a standard architecture 498 00:13:48,260 --> 00:13:49,680 of Cockroach 499 00:13:49,680 --> 00:13:50,440 or SQL Plus 500 00:13:50,440 --> 00:13:50,740 whatever 501 00:13:50,740 --> 00:13:51,300 that you've seen 502 00:13:51,300 --> 00:13:51,820 pretty commonly. 503 00:13:52,660 --> 00:13:53,060 Yeah. 504 00:13:53,240 --> 00:13:53,480 You know, 505 00:13:53,640 --> 00:13:54,780 I can't say like 506 00:13:54,780 --> 00:13:55,680 we see so much 507 00:13:55,680 --> 00:13:56,320 above the stack 508 00:13:56,320 --> 00:13:56,860 that's common. 509 00:13:57,220 --> 00:13:57,940 There seems to be 510 00:13:57,940 --> 00:13:59,140 a little bit more froth there 511 00:13:59,140 --> 00:14:00,000 but certainly, 512 00:14:00,160 --> 00:14:00,460 you know, 513 00:14:00,500 --> 00:14:00,820 Kubernetes 514 00:14:00,820 --> 00:14:01,760 as a substrate, 515 00:14:02,000 --> 00:14:03,020 this has actually 516 00:14:03,020 --> 00:14:04,240 just completely changed 517 00:14:04,240 --> 00:14:05,460 over the past four years. 518 00:14:05,880 --> 00:14:06,500 When we were getting 519 00:14:06,500 --> 00:14:07,820 started with Cockroach Labs, 520 00:14:08,360 --> 00:14:08,540 you know, 521 00:14:08,740 --> 00:14:09,620 I think Kubernetes 522 00:14:09,620 --> 00:14:10,340 was announced 523 00:14:10,340 --> 00:14:11,160 just right around 524 00:14:11,160 --> 00:14:11,740 that time 525 00:14:11,740 --> 00:14:12,880 or just slightly afterwards 526 00:14:12,880 --> 00:14:13,560 and it's like, 527 00:14:13,620 --> 00:14:13,700 oh, 528 00:14:13,720 --> 00:14:14,520 let's see how this goes 529 00:14:14,520 --> 00:14:15,060 and now it's just 530 00:14:15,060 --> 00:14:16,060 every conversation 531 00:14:16,060 --> 00:14:17,520 and every customer 532 00:14:17,520 --> 00:14:18,580 is asking about this 533 00:14:18,580 --> 00:14:20,240 and this goes from, 534 00:14:20,280 --> 00:14:20,600 you know, 535 00:14:20,640 --> 00:14:21,460 the small startups 536 00:14:21,460 --> 00:14:22,860 all the way up to, 537 00:14:23,060 --> 00:14:23,180 you know, 538 00:14:23,220 --> 00:14:23,900 the big fortune 539 00:14:23,900 --> 00:14:24,920 500 companies 540 00:14:24,920 --> 00:14:27,060 that are using us 541 00:14:27,060 --> 00:14:27,980 and experimenting 542 00:14:27,980 --> 00:14:28,360 with us. 543 00:14:28,480 --> 00:14:29,240 They all have 544 00:14:29,240 --> 00:14:29,980 some kind of Kubernetes 545 00:14:29,980 --> 00:14:31,200 mandate it feels like. 546 00:14:31,660 --> 00:14:32,320 So that's really 547 00:14:32,320 --> 00:14:33,040 the common thread 548 00:14:33,040 --> 00:14:34,140 we see going on there. 549 00:14:34,720 --> 00:14:35,600 But as far as like, 550 00:14:35,640 --> 00:14:35,820 you know, 551 00:14:35,840 --> 00:14:36,480 like kind of the other 552 00:14:36,480 --> 00:14:37,300 parts of the stack, 553 00:14:37,660 --> 00:14:37,780 you know, 554 00:14:37,780 --> 00:14:38,480 people using us 555 00:14:38,480 --> 00:14:39,100 from JavaScript, 556 00:14:39,380 --> 00:14:39,880 from Java, 557 00:14:40,060 --> 00:14:40,640 from Go, 558 00:14:41,180 --> 00:14:41,960 from Python, 559 00:14:42,480 --> 00:14:42,680 you know, 560 00:14:42,680 --> 00:14:43,100 it's just, 561 00:14:43,620 --> 00:14:44,880 it's quite a bit 562 00:14:44,880 --> 00:14:45,860 more diffuse there. 563 00:14:46,700 --> 00:14:46,920 All right. 564 00:14:47,340 --> 00:14:47,800 And then, 565 00:14:47,940 --> 00:14:48,200 I mean, 566 00:14:48,200 --> 00:14:48,940 you touched on 567 00:14:48,940 --> 00:14:49,760 the open source nature 568 00:14:49,760 --> 00:14:51,440 of Cockroach 569 00:14:51,440 --> 00:14:52,400 and a lot of other, 570 00:14:52,500 --> 00:14:52,700 you know, 571 00:14:52,800 --> 00:14:53,100 of, 572 00:14:53,240 --> 00:14:53,360 you know, 573 00:14:53,400 --> 00:14:53,960 maybe Cockroach 574 00:14:53,960 --> 00:14:54,440 alternatives 575 00:14:54,440 --> 00:14:55,440 are open source 576 00:14:55,440 --> 00:14:55,780 as well. 577 00:14:55,880 --> 00:14:56,080 Like, 578 00:14:56,420 --> 00:14:57,560 what's the reason 579 00:14:57,560 --> 00:14:58,540 behind the open source, 580 00:14:58,640 --> 00:14:58,800 you know, 581 00:14:58,800 --> 00:14:59,500 behind the open source? 582 00:14:59,540 --> 00:14:59,920 Because it seems like 583 00:14:59,920 --> 00:15:00,860 most new databases, 584 00:15:01,020 --> 00:15:01,300 frankly, 585 00:15:01,760 --> 00:15:02,040 you know, 586 00:15:02,080 --> 00:15:02,940 or at least many 587 00:15:02,940 --> 00:15:04,120 are open source. 588 00:15:04,240 --> 00:15:04,620 And then how should 589 00:15:04,620 --> 00:15:05,400 enterprises think 590 00:15:05,400 --> 00:15:06,140 about using, 591 00:15:06,220 --> 00:15:06,940 you know, 592 00:15:06,980 --> 00:15:07,720 the free versus 593 00:15:07,720 --> 00:15:08,460 the paid versions 594 00:15:08,460 --> 00:15:09,100 of those products? 595 00:15:09,400 --> 00:15:09,660 Yeah. 596 00:15:10,320 --> 00:15:10,720 Well, 597 00:15:11,080 --> 00:15:11,740 there's kind of 598 00:15:11,740 --> 00:15:12,560 two questions there. 599 00:15:12,840 --> 00:15:13,040 So, 600 00:15:13,200 --> 00:15:13,360 I mean, 601 00:15:13,380 --> 00:15:14,160 the first question is, 602 00:15:14,200 --> 00:15:14,360 you know, 603 00:15:14,420 --> 00:15:15,080 what's the motivation 604 00:15:15,080 --> 00:15:15,600 behind, 605 00:15:15,700 --> 00:15:16,500 you know, 606 00:15:16,600 --> 00:15:16,840 choosing, 607 00:15:17,000 --> 00:15:17,680 using open source 608 00:15:17,680 --> 00:15:18,400 for these products? 609 00:15:18,820 --> 00:15:19,720 And part of that is, 610 00:15:19,820 --> 00:15:19,900 you know, 611 00:15:19,920 --> 00:15:20,560 there's been some, 612 00:15:20,620 --> 00:15:21,000 you know, 613 00:15:21,060 --> 00:15:22,880 problematic history 614 00:15:22,880 --> 00:15:23,780 behind closed source 615 00:15:23,780 --> 00:15:25,080 databases provided 616 00:15:25,080 --> 00:15:25,660 by Starps. 617 00:15:26,360 --> 00:15:26,980 Foundation DB 618 00:15:26,980 --> 00:15:27,540 was the one 619 00:15:27,540 --> 00:15:28,020 that occurred 620 00:15:28,020 --> 00:15:29,060 just right around 621 00:15:29,060 --> 00:15:29,420 the time, 622 00:15:29,640 --> 00:15:29,800 you know, 623 00:15:29,800 --> 00:15:30,440 four years ago 624 00:15:30,440 --> 00:15:31,160 that Cockroach Labs 625 00:15:31,160 --> 00:15:31,720 was getting started 626 00:15:31,720 --> 00:15:32,520 where they were 627 00:15:32,520 --> 00:15:33,120 closed source, 628 00:15:33,260 --> 00:15:34,660 kind of not quite 629 00:15:34,660 --> 00:15:35,240 new SQL. 630 00:15:35,400 --> 00:15:36,080 They weren't quite 631 00:15:36,080 --> 00:15:36,600 as advanced 632 00:15:36,600 --> 00:15:37,640 as CockroachDB 633 00:15:37,640 --> 00:15:38,220 in some ways. 634 00:15:38,660 --> 00:15:39,140 They were more 635 00:15:39,140 --> 00:15:40,620 of a key value 636 00:15:40,620 --> 00:15:42,800 store type 637 00:15:42,800 --> 00:15:43,220 no SQL, 638 00:15:43,640 --> 00:15:44,080 but they were 639 00:15:44,080 --> 00:15:44,600 closed source 640 00:15:44,600 --> 00:15:45,020 and, 641 00:15:45,120 --> 00:15:45,960 you know, 642 00:15:46,300 --> 00:15:46,840 they looked like 643 00:15:46,840 --> 00:15:47,560 they had a really 644 00:15:47,560 --> 00:15:48,200 nice architecture 645 00:15:48,200 --> 00:15:48,720 and everything 646 00:15:48,720 --> 00:15:49,500 and then they got 647 00:15:49,500 --> 00:15:50,120 bought by Apple 648 00:15:50,120 --> 00:15:51,160 and got shut down. 649 00:15:51,740 --> 00:15:51,820 And, 650 00:15:51,920 --> 00:15:52,200 you know, 651 00:15:52,220 --> 00:15:52,700 that was kind of, 652 00:15:52,800 --> 00:15:52,980 you know, 653 00:15:53,040 --> 00:15:53,460 we were paying 654 00:15:53,460 --> 00:15:54,140 attention at the time 655 00:15:54,140 --> 00:15:54,480 being like, 656 00:15:54,560 --> 00:15:54,720 ah, 657 00:15:54,780 --> 00:15:55,000 you know, 658 00:15:55,060 --> 00:15:55,600 customers are going 659 00:15:55,600 --> 00:15:56,080 to have a hard 660 00:15:56,080 --> 00:15:56,940 time adopting us 661 00:15:56,940 --> 00:15:57,140 if, 662 00:15:57,340 --> 00:15:57,500 you know, 663 00:15:57,800 --> 00:15:58,600 the rug can get 664 00:15:58,600 --> 00:15:59,660 pulled out at any time. 665 00:16:00,160 --> 00:16:00,520 So, 666 00:16:00,680 --> 00:16:01,180 that was one 667 00:16:01,180 --> 00:16:01,960 of our motivations 668 00:16:01,960 --> 00:16:03,260 for going open source 669 00:16:03,260 --> 00:16:03,640 is to, 670 00:16:03,680 --> 00:16:03,840 you know, 671 00:16:03,840 --> 00:16:04,780 provide this guarantee. 672 00:16:04,920 --> 00:16:05,180 It's like, 673 00:16:05,240 --> 00:16:05,440 you know, 674 00:16:05,440 --> 00:16:06,040 no matter what else 675 00:16:06,040 --> 00:16:06,680 happens to us, 676 00:16:06,720 --> 00:16:07,240 we have bought, 677 00:16:07,340 --> 00:16:07,600 whatever, 678 00:16:08,100 --> 00:16:09,520 we go under, 679 00:16:09,660 --> 00:16:09,940 you know, 680 00:16:09,980 --> 00:16:10,500 that this is going 681 00:16:10,500 --> 00:16:11,040 to be there 682 00:16:11,040 --> 00:16:12,220 and you have 683 00:16:12,220 --> 00:16:13,460 that fail safe there. 684 00:16:13,800 --> 00:16:14,420 Another aspect 685 00:16:14,420 --> 00:16:15,920 is just the marketing, 686 00:16:16,040 --> 00:16:16,600 the go-to-market 687 00:16:16,600 --> 00:16:17,600 of using open source. 688 00:16:18,340 --> 00:16:19,020 There's a kind of 689 00:16:19,020 --> 00:16:20,060 heavyweight go-to-market 690 00:16:20,060 --> 00:16:21,020 asking people like, 691 00:16:21,200 --> 00:16:21,260 hey, 692 00:16:21,340 --> 00:16:22,260 try my closed source 693 00:16:22,260 --> 00:16:23,200 database software 694 00:16:23,200 --> 00:16:24,060 and, 695 00:16:24,240 --> 00:16:24,680 you know, 696 00:16:25,140 --> 00:16:25,860 you usually have 697 00:16:25,860 --> 00:16:27,080 a kind of heavy sales 698 00:16:27,080 --> 00:16:28,160 effort involved there 699 00:16:28,160 --> 00:16:28,660 where, 700 00:16:28,760 --> 00:16:28,900 you know, 701 00:16:28,940 --> 00:16:29,200 it's like, 702 00:16:29,260 --> 00:16:29,900 you want to get 703 00:16:29,900 --> 00:16:30,920 the access to the binary 704 00:16:30,920 --> 00:16:31,960 for trial purposes, 705 00:16:32,420 --> 00:16:33,100 three-day trial 706 00:16:33,100 --> 00:16:33,440 and whatnot. 707 00:16:34,300 --> 00:16:35,440 What open source does 708 00:16:35,440 --> 00:16:35,940 is it allows you 709 00:16:35,940 --> 00:16:36,340 to kind of go 710 00:16:36,340 --> 00:16:37,040 in the back door. 711 00:16:37,500 --> 00:16:38,240 Developers can just 712 00:16:38,240 --> 00:16:39,080 download your software, 713 00:16:39,280 --> 00:16:40,160 start trying it, 714 00:16:40,220 --> 00:16:40,860 kicking the tires 715 00:16:40,860 --> 00:16:41,200 and whatnot. 716 00:16:41,560 --> 00:16:42,180 They can even peek 717 00:16:42,180 --> 00:16:42,680 under the hood 718 00:16:42,680 --> 00:16:44,220 to see if the claims 719 00:16:44,220 --> 00:16:44,580 you're making 720 00:16:44,580 --> 00:16:45,360 on your website 721 00:16:45,360 --> 00:16:46,280 are valid. 722 00:16:46,800 --> 00:16:46,920 You know, 723 00:16:46,960 --> 00:16:48,020 you hear someone says, 724 00:16:48,180 --> 00:16:48,300 oh, 725 00:16:48,340 --> 00:16:49,460 this is absolutely reliable 726 00:16:49,460 --> 00:16:51,900 and then you actually 727 00:16:51,900 --> 00:16:52,520 peek under the hood 728 00:16:52,520 --> 00:16:53,880 and you see what 729 00:16:53,880 --> 00:16:54,900 their coding practices are 730 00:16:54,900 --> 00:16:55,300 and see what 731 00:16:55,300 --> 00:16:56,460 their testing practices are 732 00:16:56,460 --> 00:16:56,660 and, 733 00:16:56,840 --> 00:16:56,980 you know, 734 00:16:57,040 --> 00:16:57,460 kind of, 735 00:16:57,720 --> 00:16:57,880 you know, 736 00:16:57,900 --> 00:16:59,100 your eyes can open wide 737 00:16:59,100 --> 00:16:59,520 and be like, 738 00:16:59,620 --> 00:16:59,760 oh, 739 00:16:59,860 --> 00:17:00,920 no way would I trust that. 740 00:17:01,280 --> 00:17:01,740 And what we do 741 00:17:01,740 --> 00:17:02,460 is all in the open. 742 00:17:02,980 --> 00:17:03,640 People can, 743 00:17:03,700 --> 00:17:03,880 you know, 744 00:17:03,880 --> 00:17:04,500 follow along. 745 00:17:04,880 --> 00:17:05,560 Not sure if they do, 746 00:17:05,660 --> 00:17:06,360 but that's part of the, 747 00:17:06,400 --> 00:17:07,120 you know, 748 00:17:07,460 --> 00:17:09,360 the rationale behind it. 749 00:17:09,580 --> 00:17:10,900 Kind of keeps us honest 750 00:17:10,900 --> 00:17:11,500 to have that 751 00:17:11,500 --> 00:17:12,640 openness baked 752 00:17:12,640 --> 00:17:13,340 into what we do. 753 00:17:13,760 --> 00:17:13,900 Okay. 754 00:17:13,980 --> 00:17:14,800 And then the second part 755 00:17:14,800 --> 00:17:15,260 of the question, 756 00:17:15,360 --> 00:17:15,480 Peter, 757 00:17:15,520 --> 00:17:16,640 was how should enterprises 758 00:17:16,640 --> 00:17:18,120 think about using 759 00:17:18,120 --> 00:17:19,240 or I guess any users 760 00:17:19,240 --> 00:17:19,880 think about using 761 00:17:19,880 --> 00:17:20,720 the free versus 762 00:17:20,720 --> 00:17:21,780 paid version 763 00:17:21,780 --> 00:17:23,600 of an open source product? 764 00:17:24,580 --> 00:17:24,980 Yeah. 765 00:17:25,500 --> 00:17:25,860 So, 766 00:17:26,080 --> 00:17:26,340 I mean, 767 00:17:26,360 --> 00:17:27,080 one of the main reasons 768 00:17:27,080 --> 00:17:28,700 that people use 769 00:17:28,700 --> 00:17:29,380 our paid version 770 00:17:29,380 --> 00:17:30,280 is we actually put 771 00:17:30,280 --> 00:17:31,940 features into 772 00:17:31,940 --> 00:17:32,880 that enterprise version 773 00:17:32,880 --> 00:17:33,900 that you want to have 774 00:17:33,900 --> 00:17:34,340 if you're running 775 00:17:34,340 --> 00:17:34,860 a production. 776 00:17:35,300 --> 00:17:35,920 And we kind of 777 00:17:35,920 --> 00:17:36,860 try to draw the line 778 00:17:36,860 --> 00:17:38,880 between our free version 779 00:17:38,880 --> 00:17:39,520 and our paid version 780 00:17:39,520 --> 00:17:40,500 based on things that, 781 00:17:40,580 --> 00:17:40,800 you know, 782 00:17:40,940 --> 00:17:41,580 a small startup 783 00:17:41,580 --> 00:17:42,400 might not necessarily 784 00:17:42,400 --> 00:17:43,040 need when they're 785 00:17:43,040 --> 00:17:43,740 just getting started, 786 00:17:43,940 --> 00:17:44,780 but you almost certainly 787 00:17:44,780 --> 00:17:45,480 want when you're getting 788 00:17:45,480 --> 00:17:46,420 to any sort of scale. 789 00:17:47,060 --> 00:17:48,120 Most obvious example 790 00:17:48,120 --> 00:17:48,620 of that is 791 00:17:48,620 --> 00:17:49,440 Backup and Restore 792 00:17:49,440 --> 00:17:50,660 is on our paid version. 793 00:17:51,160 --> 00:17:51,380 And, 794 00:17:51,440 --> 00:17:51,880 you know, 795 00:17:51,880 --> 00:17:52,820 like any big enterprise 796 00:17:52,820 --> 00:17:53,480 that's using 797 00:17:53,480 --> 00:17:54,200 a database product 798 00:17:54,200 --> 00:17:55,140 is probably going to want 799 00:17:55,140 --> 00:17:55,960 Backup and Restore. 800 00:17:56,500 --> 00:17:58,000 But above and beyond that, 801 00:17:58,260 --> 00:17:58,520 you know, 802 00:17:58,560 --> 00:17:59,500 there's usually, 803 00:17:59,660 --> 00:17:59,920 you know, 804 00:17:59,980 --> 00:18:00,780 support contracts 805 00:18:00,780 --> 00:18:01,320 that come along 806 00:18:01,320 --> 00:18:02,280 with our enterprise version. 807 00:18:02,800 --> 00:18:03,800 And that's also 808 00:18:03,800 --> 00:18:04,800 kind of a critical need 809 00:18:04,800 --> 00:18:05,580 for any enterprise. 810 00:18:05,740 --> 00:18:06,260 If you're putting, 811 00:18:06,460 --> 00:18:06,580 you know, 812 00:18:06,620 --> 00:18:07,180 we want people 813 00:18:07,180 --> 00:18:08,280 to put their crown jewels 814 00:18:08,280 --> 00:18:08,880 with their data 815 00:18:08,880 --> 00:18:10,080 into CockroachDB, 816 00:18:10,740 --> 00:18:11,820 it's kind of almost 817 00:18:11,820 --> 00:18:12,660 a fiduciary duty 818 00:18:12,660 --> 00:18:13,680 to have a support contract 819 00:18:13,680 --> 00:18:14,360 at that point 820 00:18:14,360 --> 00:18:15,000 with the company 821 00:18:15,000 --> 00:18:16,040 providing that software. 822 00:18:16,740 --> 00:18:17,020 And then, 823 00:18:17,120 --> 00:18:18,220 and then finally, 824 00:18:18,320 --> 00:18:18,600 Peter, 825 00:18:19,260 --> 00:18:19,620 I mean, 826 00:18:19,640 --> 00:18:21,000 what does your ideal future 827 00:18:21,000 --> 00:18:21,420 look like 828 00:18:21,420 --> 00:18:22,320 with regard to 829 00:18:22,320 --> 00:18:24,520 SQL and database technology 830 00:18:24,520 --> 00:18:24,920 in general, 831 00:18:25,000 --> 00:18:25,120 right? 832 00:18:25,120 --> 00:18:25,260 I mean, 833 00:18:25,280 --> 00:18:25,860 that's in terms 834 00:18:25,860 --> 00:18:26,460 of capabilities, 835 00:18:26,600 --> 00:18:26,940 I guess, 836 00:18:27,060 --> 00:18:27,920 but also in terms 837 00:18:27,920 --> 00:18:28,820 of how you would 838 00:18:28,820 --> 00:18:29,500 like to see 839 00:18:29,500 --> 00:18:31,100 organizations and users, 840 00:18:31,300 --> 00:18:31,780 you know, 841 00:18:31,820 --> 00:18:32,340 architecting 842 00:18:32,340 --> 00:18:32,640 and building 843 00:18:32,640 --> 00:18:33,280 new applications 844 00:18:33,280 --> 00:18:33,960 to take advantage 845 00:18:33,960 --> 00:18:35,180 of what's coming 846 00:18:35,180 --> 00:18:35,640 down the pike. 847 00:18:37,380 --> 00:18:37,740 Yeah, 848 00:18:37,900 --> 00:18:38,120 I mean, 849 00:18:38,160 --> 00:18:38,880 the thing I'm excited 850 00:18:38,880 --> 00:18:39,600 about is the, 851 00:18:39,660 --> 00:18:39,880 you know, 852 00:18:39,940 --> 00:18:40,760 this evolving 853 00:18:40,760 --> 00:18:43,600 landscape of applications 854 00:18:43,600 --> 00:18:44,120 that can work 855 00:18:44,120 --> 00:18:44,840 in a global manner 856 00:18:44,840 --> 00:18:46,200 and have geodistributed 857 00:18:46,200 --> 00:18:46,880 applications 858 00:18:46,880 --> 00:18:48,420 that can take advantage 859 00:18:48,420 --> 00:18:49,760 of a geodistributed database. 860 00:18:50,320 --> 00:18:50,980 And there's going 861 00:18:50,980 --> 00:18:51,720 to be enhancements 862 00:18:51,720 --> 00:18:53,520 to the kind 863 00:18:53,520 --> 00:18:54,620 of the SQL abstractions 864 00:18:54,620 --> 00:18:55,840 to make this easier to do. 865 00:18:56,080 --> 00:18:56,420 And I'd say 866 00:18:56,420 --> 00:18:57,020 we're still kind 867 00:18:57,020 --> 00:18:57,680 in the early days 868 00:18:57,680 --> 00:18:58,020 of this. 869 00:18:58,500 --> 00:18:58,940 CockroachDB 870 00:18:58,940 --> 00:18:59,760 provides a lot 871 00:18:59,760 --> 00:19:00,220 of mechanism 872 00:19:00,220 --> 00:19:00,820 for doing 873 00:19:00,820 --> 00:19:02,340 geodistributed applications. 874 00:19:03,140 --> 00:19:03,980 And yet those mechanisms 875 00:19:03,980 --> 00:19:04,840 are a little bit, 876 00:19:04,860 --> 00:19:05,180 you know, 877 00:19:05,340 --> 00:19:06,260 rough around the edges. 878 00:19:06,600 --> 00:19:06,620 So, 879 00:19:06,800 --> 00:19:06,920 I mean, 880 00:19:06,980 --> 00:19:07,860 the area I'm excited 881 00:19:07,860 --> 00:19:08,540 to see innovation 882 00:19:08,540 --> 00:19:10,380 in is improving 883 00:19:10,380 --> 00:19:11,940 those SQL abstractions, 884 00:19:12,180 --> 00:19:12,880 improving how, 885 00:19:13,120 --> 00:19:13,280 you know, 886 00:19:13,300 --> 00:19:14,200 that functionality works. 887 00:19:14,260 --> 00:19:15,320 So it's just crystal clear 888 00:19:15,320 --> 00:19:16,040 and easy 889 00:19:16,040 --> 00:19:17,200 for developers to use. 890 00:19:17,480 --> 00:19:17,900 And then also 891 00:19:17,900 --> 00:19:18,800 seeing this evolution 892 00:19:18,800 --> 00:19:19,880 of the, 893 00:19:20,140 --> 00:19:20,300 you know, 894 00:19:20,320 --> 00:19:21,400 kind of a global app 895 00:19:21,400 --> 00:19:22,780 tier come into play. 896 00:19:23,220 --> 00:19:23,420 You can, 897 00:19:23,900 --> 00:19:24,700 so easy 898 00:19:24,700 --> 00:19:25,900 to get a VM 899 00:19:25,900 --> 00:19:27,000 running in a data center, 900 00:19:27,020 --> 00:19:27,920 but actually getting 901 00:19:27,920 --> 00:19:28,820 an application running 902 00:19:28,820 --> 00:19:29,600 is quite a bit harder 903 00:19:29,600 --> 00:19:30,240 than trying to get 904 00:19:30,240 --> 00:19:30,940 a geodistributed 905 00:19:30,940 --> 00:19:31,680 application running 906 00:19:31,680 --> 00:19:32,860 is even harder still. 907 00:19:33,060 --> 00:19:33,940 And all that stuff 908 00:19:33,940 --> 00:19:34,680 is going to be addressed 909 00:19:34,680 --> 00:19:35,220 over time. 910 00:19:35,340 --> 00:19:36,240 We're seeing progress 911 00:19:36,240 --> 00:19:36,900 in these areas. 912 00:19:37,420 --> 00:19:37,580 You know, 913 00:19:37,780 --> 00:19:38,440 before it was starting 914 00:19:38,440 --> 00:19:38,860 a VM, 915 00:19:39,320 --> 00:19:40,000 now it's like 916 00:19:40,000 --> 00:19:40,740 usually you get 917 00:19:40,740 --> 00:19:41,600 a Kubernetes cluster 918 00:19:41,600 --> 00:19:42,020 running, 919 00:19:42,380 --> 00:19:43,480 but getting a multi-region 920 00:19:43,480 --> 00:19:44,280 Kubernetes cluster 921 00:19:44,280 --> 00:19:45,020 is still hard. 922 00:19:45,540 --> 00:19:46,260 I think these things 923 00:19:46,260 --> 00:19:46,980 are all going to be 924 00:19:46,980 --> 00:19:48,060 addressed in the coming years. 925 00:19:48,420 --> 00:19:49,020 And we're going to get 926 00:19:49,020 --> 00:19:49,520 to this point 927 00:19:49,520 --> 00:19:49,800 where, 928 00:19:49,900 --> 00:19:50,080 you know, 929 00:19:50,100 --> 00:19:50,820 if I want to, 930 00:19:50,820 --> 00:19:51,080 you know, 931 00:19:51,080 --> 00:19:52,780 spin up a global application, 932 00:19:52,860 --> 00:19:53,560 I'll be able to do that 933 00:19:53,560 --> 00:19:54,020 in minutes 934 00:19:54,020 --> 00:19:54,960 and deploy it in minutes. 935 00:19:55,760 --> 00:19:55,940 You know, 936 00:19:56,060 --> 00:19:56,580 that's still, 937 00:19:56,700 --> 00:19:56,880 you know, 938 00:19:56,880 --> 00:19:57,560 kind of a far off 939 00:19:57,560 --> 00:19:58,160 pie in the sky, 940 00:19:58,280 --> 00:19:59,240 but I think 941 00:19:59,240 --> 00:19:59,800 we'll get there. 942 00:20:00,800 --> 00:20:01,320 All right. 943 00:20:01,360 --> 00:20:01,600 Great. 944 00:20:01,920 --> 00:20:02,360 And with that, 945 00:20:02,740 --> 00:20:03,520 we are out of time. 946 00:20:03,600 --> 00:20:04,120 Thanks a lot, Peter. 947 00:20:05,280 --> 00:20:05,860 All right. 948 00:20:06,080 --> 00:20:06,780 Well, thank you, Derek. 949 00:20:07,640 --> 00:20:08,520 There you have it. 950 00:20:08,560 --> 00:20:10,600 I hope it was 20 minutes 951 00:20:10,600 --> 00:20:11,140 well spent. 952 00:20:11,700 --> 00:20:12,540 For more information 953 00:20:12,540 --> 00:20:13,620 about Cockroach Labs, 954 00:20:13,720 --> 00:20:14,160 the company, 955 00:20:14,380 --> 00:20:15,220 and CockroachDB, 956 00:20:15,360 --> 00:20:15,860 the product, 957 00:20:16,100 --> 00:20:16,860 you can visit 958 00:20:16,860 --> 00:20:18,540 CockroachLabs.com 959 00:20:18,540 --> 00:20:19,520 or you can dive 960 00:20:19,520 --> 00:20:20,120 a little deeper 961 00:20:20,120 --> 00:20:21,040 into the project 962 00:20:21,040 --> 00:20:22,720 at GitHub.com 963 00:20:22,720 --> 00:20:24,180 slash CockroachDB. 964 00:20:24,820 --> 00:20:25,840 For more information 965 00:20:25,840 --> 00:20:27,120 on our broad suite 966 00:20:27,120 --> 00:20:27,860 of products 967 00:20:27,860 --> 00:20:28,440 and services 968 00:20:28,440 --> 00:20:28,940 for building 969 00:20:28,940 --> 00:20:29,420 and managing 970 00:20:29,420 --> 00:20:30,480 cloud-native applications, 971 00:20:30,940 --> 00:20:32,060 as well as transforming 972 00:20:32,060 --> 00:20:33,040 how you build software, 973 00:20:33,560 --> 00:20:35,260 check out Pivotal.io. 974 00:20:35,780 --> 00:20:36,360 And if you like 975 00:20:36,360 --> 00:20:36,960 this podcast 976 00:20:36,960 --> 00:20:37,840 and want to keep up 977 00:20:37,840 --> 00:20:38,780 with all things 978 00:20:38,780 --> 00:20:39,700 quote-unquote 979 00:20:39,700 --> 00:20:40,800 digital transformation, 980 00:20:41,340 --> 00:20:42,360 including some insightful 981 00:20:42,360 --> 00:20:43,480 posts from experts, 982 00:20:43,660 --> 00:20:44,900 some guides and explainers 983 00:20:44,900 --> 00:20:46,080 on cloud-native technology, 984 00:20:46,480 --> 00:20:47,360 and even highlights 985 00:20:47,360 --> 00:20:48,260 from these podcasts, 986 00:20:48,700 --> 00:20:49,340 you can visit 987 00:20:49,340 --> 00:20:50,460 Pivotal.io 988 00:20:50,460 --> 00:20:51,580 slash intersect 989 00:20:51,580 --> 00:20:53,420 as in the intersection 990 00:20:53,420 --> 00:20:54,220 of business 991 00:20:54,220 --> 00:20:55,000 and technology. 992 00:20:55,480 --> 00:20:55,780 That's 993 00:20:55,780 --> 00:20:57,180 Pivotal.io 994 00:20:57,180 --> 00:20:58,420 slash intersect. 995 00:20:59,140 --> 00:21:00,100 Thanks for listening.