97091c8c-ced7-4b87-a0c7-26b8a5336808-0 If you do have any questions, please enter them into the 97091c8c-ced7-4b87-a0c7-26b8a5336808-1 Q&A box within the Zoom window, and we'll try to get to 97091c8c-ced7-4b87-a0c7-26b8a5336808-2 as many as possible during the Q and A portion of the 97091c8c-ced7-4b87-a0c7-26b8a5336808-3 presentation. 1dad90a8-c37e-4762-a045-080fb89c8312-0 So now let's jump into webinar. e8be59a7-878b-483d-b66e-bbd9cb489ee3-0 This presentation. 4a0c4cf8-0b5c-4431-8c4c-6ebc81066e0f-0 We have senior solutions Architect Andrew Mills here with 4a0c4cf8-0b5c-4431-8c4c-6ebc81066e0f-1 us today to discuss Apache Kafka and avoid using your streaming 4a0c4cf8-0b5c-4431-8c4c-6ebc81066e0f-2 platform as a database. 61b1e6e6-73ca-465b-8ab0-348e97791f78-0 All right, Andrew, I'll pass it over to you to get things 61b1e6e6-73ca-465b-8ab0-348e97791f78-1 started. 9c09ab52-228d-4932-af88-588afd0c95ce-0 Thanks, Steven. 4853f7d4-ed71-4d04-9806-b3d0e0b01924-0 Thank you so much. 3ea8912a-dbcd-4521-877d-b2ba10d7cd12-0 As he mentioned, my name is Andrew Mills and I'm a senior 3ea8912a-dbcd-4521-877d-b2ba10d7cd12-1 solution architect here at Instacluster, which is now part 3ea8912a-dbcd-4521-877d-b2ba10d7cd12-2 of spot by Net app. 9742ca65-12e0-429c-86d0-125578438e2a-0 So I know that's thebusiness intro, but really, who 9742ca65-12e0-429c-86d0-125578438e2a-1 am I? 7d4a251c-dcfb-4546-b3fb-b7ced4acd0af-0 So I am a husband and a father and a scallywag sometimes, 7d4a251c-dcfb-4546-b3fb-b7ced4acd0af-1 right? 78be0fa7-2277-425c-a1e8-921e433198ef-0 More importantly, a thinker. ca096352-3992-468a-9e76-1325f0f18e51-0 And most importantly for this conversation, I'm a Kafka ca096352-3992-468a-9e76-1325f0f18e51-1 expert, right? 61b689eb-d59d-47ff-b607-2ac20845545e-0 So I've been working with Apache Kafka since O .11. a952b5ec-ce7f-48a1-83dd-28559bec9a67-0 I've built various data pipelines from pretty big ones a952b5ec-ce7f-48a1-83dd-28559bec9a67-1 to pretty small ones, and I'veused quite a few tools to a952b5ec-ce7f-48a1-83dd-28559bec9a67-2 deploy Apache Kafka. 3c28f38d-2698-456f-8b60-d308bf2933e8-0 Some of them specifically are Data stacks, OPS Center, the 3c28f38d-2698-456f-8b60-d308bf2933e8-1 Hortonworks Data Platform, which now I guess is the Cloudera Data 3c28f38d-2698-456f-8b60-d308bf2933e8-2 platform, kind of. 1e60c127-e82e-421d-b6a2-6ca84c3b97f6-0 I've used pure open source Kafka, Strimzee, Confluent 1e60c127-e82e-421d-b6a2-6ca84c3b97f6-1 Cloud, Confluent Community Edition, and also Instacluster, 1e60c127-e82e-421d-b6a2-6ca84c3b97f6-2 which is a really good offering. f3b73344-d921-4a8d-9b0a-8b9e76f0ad63-0 I'm passionate about the event driven architecture and the open f3b73344-d921-4a8d-9b0a-8b9e76f0ad63-1 source community as a whole and have been since early 2015. 02d9583b-6437-4419-8dd4-7c5e9a44d69b-0 So why are we here? 77c6a715-04d3-43fc-888c-672a9386d060-0 So we're here to talk about event driven architecture and 77c6a715-04d3-43fc-888c-672a9386d060-1 how I believe it's essential to a healthy data-driven ecosystem. b8d9023d-1971-4dd6-b7f3-276fb7fc30a8-0 Most organizations, they're already using event driven b8d9023d-1971-4dd6-b7f3-276fb7fc30a8-1 architecture or they're moving that way and navigating from a b8d9023d-1971-4dd6-b7f3-276fb7fc30a8-2 traditional like data storage and retrieval process. d424c91d-7387-4918-94e6-f31a38f5078a-0 Using relational databases is complicated when you're moving d424c91d-7387-4918-94e6-f31a38f5078a-1 to EDA. eaf7eaf4-710f-4b0c-a634-beb887d0fde3-0 So the purpose of this conversation is to discuss eaf7eaf4-710f-4b0c-a634-beb887d0fde3-1 Apache Kafka, the role it plays in this transformation, and eaf7eaf4-710f-4b0c-a634-beb887d0fde3-2 specifically, will it replace my database? 79322c3a-d9a1-4228-8fad-9c2f06338a5f-0 We're going to use a pretty simple use case to help 79322c3a-d9a1-4228-8fad-9c2f06338a5f-1 understand where Kafka can fit. f47bcf0e-24b7-4af3-ac02-5ed8fd888dcc-0 And this is not prescriptive. f3b77248-ac06-419e-a2a7-3f86a7432f70-0 There's,a billion different ways to use Kafka. 102bbb1b-e314-47dd-b14a-460d16339241-0 And for some of you, this,use case may be kind of old, but 102bbb1b-e314-47dd-b14a-460d16339241-1 for a lot of the senior level architects or decision makers, 102bbb1b-e314-47dd-b14a-460d16339241-2 this is, this is a lot of their genesis, right? b04f891f-cf98-422c-a0ff-2b39a59fc920-0 This is where they came from because we didn't have platforms b04f891f-cf98-422c-a0ff-2b39a59fc920-1 like Amazon or Wix or some of these other tools where you can b04f891f-cf98-422c-a0ff-2b39a59fc920-2 easily sell things, right? 508bc7d2-b262-4b9b-bfd0-0e4319395564-0 And so that's going to be our, that's going to be our simple 508bc7d2-b262-4b9b-bfd0-0e4319395564-1 use case, right? 5de776a0-f668-4c20-8d08-a212c66cf2e9-0 The simple use case here is let's sell some stuff, OK? 8aeb1800-3d69-4659-bd1d-b5b7b5bf57cd-0 And so in this use case, you're going to have your users, and 8aeb1800-3d69-4659-bd1d-b5b7b5bf57cd-1 those users are going to hit a website, and the website will 8aeb1800-3d69-4659-bd1d-b5b7b5bf57cd-2 talk to an API for a variety of things. 628d2c64-d511-4df8-8293-08353ca4f353-0 But in this case, we're going to have this user creating an 628d2c64-d511-4df8-8293-08353ca4f353-1 account and placing an order. eb972a7e-7472-4e6c-80e0-b76513728925-0 And so that API is then going to interface with the database. 054488c6-2612-4978-aaf0-7e4a43c28ea3-0 Again, traditionally this was a relational database, something 054488c6-2612-4978-aaf0-7e4a43c28ea3-1 like SQL Server or Postgres, and everything would get stored 054488c6-2612-4978-aaf0-7e4a43c28ea3-2 there as one big monolithic database. 1443cd99-6a93-4f17-a92b-3478551c7757-0 Well, then you would have order processing, right? 6eeb8676-3098-4e2d-bdf6-5ef9e4240de1-0 So the order processing engine would pick out the order 6eeb8676-3098-4e2d-bdf6-5ef9e4240de1-1 details, customer details, you know, shipping information, make 6eeb8676-3098-4e2d-bdf6-5ef9e4240de1-2 sure they paid so forth and so on. ff2c7e5d-3233-47e0-ad97-fdc7553c7a51-0 And,then they do the order processing. 70404b7e-ec16-4f37-8707-e1203c676990-0 And again, now it all happened from the relational database. 3f16657e-f69d-4c9f-a4ed-592b7ca669ce-0 From that same database you could build reports. 743350d8-219d-4e22-b9b7-fbb5f6704145-0 Now we're talking daily reports, weekly reports, monthly, yearly 743350d8-219d-4e22-b9b7-fbb5f6704145-1 roll ups, right? ebc2ed31-74ae-495a-bc13-e2d6668e128c-0 You can do all sorts of fun stuff in that relational ebc2ed31-74ae-495a-bc13-e2d6668e128c-1 database. d098eb9b-08c4-45ad-92af-ad7a1a8a05fc-0 And you could also generate metrics, right? 73531c0e-933f-4f4e-838c-8dc2da2834bc-0 You could pull metrics hourly, real time was kind of a stretch, 73531c0e-933f-4f4e-838c-8dc2da2834bc-1 but you know, you could do them, you know, on demand refreshes or 73531c0e-933f-4f4e-838c-8dc2da2834bc-2 like hourly refreshes and some of the old school 73531c0e-933f-4f4e-838c-8dc2da2834bc-3 reporting tools, right? 2f4053f7-112e-46ff-86c2-b043d16f3e13-0 And So what inevitably happens is now you've got a great system 2f4053f7-112e-46ff-86c2-b043d16f3e13-1 set up, you're taking orders, you're processing, everything's 2f4053f7-112e-46ff-86c2-b043d16f3e13-2 working great, you're generating revenue and the business comes 2f4053f7-112e-46ff-86c2-b043d16f3e13-3 to you and they're like, Hey, we want more business, we're going 2f4053f7-112e-46ff-86c2-b043d16f3e13-4 to offer more products, right? c08845f5-9759-4f87-a450-b1b257f8476d-0 So some of the questions am I ask you is marketing comes and c08845f5-9759-4f87-a450-b1b257f8476d-1 says, Hey, how many visitors are we getting on our website? 7bbb5a00-6549-4d11-810e-0dab65f8cfcc-0 Well, back then it was pretty easy, all right? dfae75d2-dae6-41a6-8123-cd22ae6d2f64-0 You just use some JavaScript, drop in a counter and,dfae75d2-dae6-41a6-8123-cd22ae6d2f64-1 you're like, there you go. aba079ad-cd4e-4d74-a7f0-cce6a771b490-0 Now we know how many people hit the website, right? a946b239-7dff-44c4-a38d-082ae73e2a15-0 So then they come and say, well, what pages are they hitting? 8d990474-f337-4593-b388-f93888a025d6-0 Well, we got an answer for that too, right? 8dabc776-d25c-4c62-b9aa-a810cd0f1cc1-0 Just another counter, but you add it to a different page or 8dabc776-d25c-4c62-b9aa-a810cd0f1cc1-1 you add it to all your pages. 8a9f0e3b-ff13-4da0-8347-4c13f5b2af3c-0 Some people got cute and added it as an invisible counter. a982656f-9c6d-4432-8829-209c2a6516eb-0 And so then you,could see it, you know, in certain ways a982656f-9c6d-4432-8829-209c2a6516eb-1 when you knew how to enable it, right. 77db0dfe-1ab2-4442-af40-b0a1ad76f0c6-0 Those are pretty easy problems to solve and they,weren't 77db0dfe-1ab2-4442-af40-b0a1ad76f0c6-1 really foolproof either. 18a53c1b-703a-4c4a-b007-b377f152b5aa-0 Well, then they come to you with a little bit more complicated 18a53c1b-703a-4c4a-b007-b377f152b5aa-1 questions like how long do they stay on every page? 6c80d3aa-53ee-41a0-954e-5bacefe08446-0 Do they like the blue button or the light blue button? 9d19cce1-26ce-4e14-9718-26ceddb63dcc-0 Right. ae4830c4-992d-4beb-9dd6-51dd75b0d422-0 Some AB testing to figure out what's the best solution. b4f14853-d010-4568-af47-404626b6260c-0 Well, we need faster process. e4ec657a-3102-4c48-a347-121635f3a075-0 We need, we need to process things faster. 84f8a6d4-1899-4a8c-8115-894083e7b3a7-0 Can they see the orders in real time, right. f34f2580-b4a6-4ca4-832b-7b1814d6177d-0 I need to show real time leadership leaderboards for my f34f2580-b4a6-4ca4-832b-7b1814d6177d-1 sales reps Are they clicking my emails right. 11cc45d9-2adb-4b4f-bcce-9d685f2fc5f5-0 So you get into these different complicated business questions 11cc45d9-2adb-4b4f-bcce-9d685f2fc5f5-1 and your infrastructure in the way that you design your 11cc45d9-2adb-4b4f-bcce-9d685f2fc5f5-2 application andreally your business, it can be a challenge 11cc45d9-2adb-4b4f-bcce-9d685f2fc5f5-3 to solve these problems using that traditional relational 11cc45d9-2adb-4b4f-bcce-9d685f2fc5f5-4 database. c278ad8d-ca00-4f47-98cd-6b96ad9c56c3-0 So how can you solve it? 997ad7d1-c9ad-4c5b-9f0a-e228d1dcb800-0 So once one solution is you kind of split up your databases. b9a8aa79-a0d1-4c08-b996-64a8ea6fdaf7-0 So as you can see here that what used to be database is now the b9a8aa79-a0d1-4c08-b996-64a8ea6fdaf7-1 orders database. 523a5590-45ce-43a5-85ef-f982982548fe-0 OK, so your API then as the customer information gets saved, 523a5590-45ce-43a5-85ef-f982982548fe-1 it's going to go into a customer's database. eb560f16-beb1-4036-aa02-c05f6f232666-0 Well, now you might have a click stream database, right? afe1b072-4acf-40bf-8beb-cee10512c926-0 You,want to do click streaming data from your website afe1b072-4acf-40bf-8beb-cee10512c926-1 or emails and you start kind of sharding off the databases, the afe1b072-4acf-40bf-8beb-cee10512c926-2 OR the,operations into different databases. 0c4a5b7d-719e-4d7c-b5d1-b1c5a1c73aa8-0 And that creates a problem because now you have a lot of 0c4a5b7d-719e-4d7c-b5d1-b1c5a1c73aa8-1 inner communication and that's fine. 31b98410-f509-4eff-a0bd-aea5f4cb69e2-0 But the,tight coupling of the reports and metrics and 31b98410-f509-4eff-a0bd-aea5f4cb69e2-1 processing with a lot of different databases can present 31b98410-f509-4eff-a0bd-aea5f4cb69e2-2 problems because the moment there's an issue or a slow down 31b98410-f509-4eff-a0bd-aea5f4cb69e2-3 in let's say your customer database, well, now maybe your 31b98410-f509-4eff-a0bd-aea5f4cb69e2-4 website's slow, they can't log in or they can't place orders or 31b98410-f509-4eff-a0bd-aea5f4cb69e2-5 they can't create accounts, right? 29b0ead1-4727-4d2a-9d85-621f1946774e-0 Or there was an outage and because of the tight coupling, 29b0ead1-4727-4d2a-9d85-621f1946774e-1 you couldn't do anything. 193ecc0f-cd70-49eb-8ca5-66ed1f60f84e-0 The order processing reports weren't showing up. 71572f2f-333b-49fa-a77b-8aa4911b80ec-0 No one can use the website, right? d34350c0-2e39-41fe-83b0-683c8bb7f2de-0 So you,get into this tight coupling situation, you're d34350c0-2e39-41fe-83b0-683c8bb7f2de-1 missing e-mail data because you had an outage, right? d90142b2-db9e-424e-a046-e151785ee9ea-0 So now all of a sudden the click stream events are gone, right? 2795bf59-5b47-45d5-bce5-e7d9c4bc3c9d-0 Reports take forever to load because you know you've got a 2795bf59-5b47-45d5-bce5-e7d9c4bc3c9d-1 giant one or two database tables where you put in additional 2795bf59-5b47-45d5-bce5-e7d9c4bc3c9d-2 indices on them. b8de9b16-d36e-442e-bbc8-37857d5980dd-0 You've done various things, but now it's got to actually cross b8de9b16-d36e-442e-bbc8-37857d5980dd-1 reference three or four or five tables and maybe even across one b8de9b16-d36e-442e-bbc8-37857d5980dd-2 or two different databases. b70f65f4-e941-4273-95fe-0595775d7de8-0 And so things start taking forever to solve. 48614899-7902-461e-b388-a926e1c1fd3b-0 So the question is the next phase, where do we go from here? 41e608fa-805c-4684-98b7-c8938915a352-0 How do we, how do we solve this problem? 8f141864-e736-430d-821f-5031fa1c1892-0 And so maybe your developers get that task and they look out on 8f141864-e736-430d-821f-5031fa1c1892-1 the market and say, you know what cough is fast, this new 8f141864-e736-430d-821f-5031fa1c1892-2 this new tool, it's scalable, it's highly available, right? 65112781-c464-451e-a5cc-9bc3569ba60d-0 And these are all things that you want because you know your 65112781-c464-451e-a5cc-9bc3569ba60d-1 company's need speed, it's going to grow and you don't 65112781-c464-451e-a5cc-9bc3569ba60d-2 want to suffer from a catastrophic outage like you had 65112781-c464-451e-a5cc-9bc3569ba60d-3 before. a90386cc-70ed-40d5-8a2c-32c11e0b751f-0 And it's new and shiny. afbdaa48-2d42-42a7-8b68-320fc0c18da2-0 And let's be honest, a lot of developers, they like to learn afbdaa48-2d42-42a7-8b68-320fc0c18da2-1 new things. e93342d6-ba0b-4a69-b824-906a43c66b59-0 And it's part of the why, part of the reason that a lot of us e93342d6-ba0b-4a69-b824-906a43c66b59-1 are in the industry is because we have a love for learning. d9bee3ce-8039-43f4-ac58-f404feaf4a73-0 So in To Save the Day, Apache Kafka, right? 2e830f1d-06ca-457b-87f3-727351e464ca-0 It's the superhero. 7f64428b-b27b-4f88-aa07-0b5aae875eb9-0 And so what's what we've seen people do in the past is they 7f64428b-b27b-4f88-aa07-0b5aae875eb9-1 kind of just try to drop it in, right? 76476435-e1c0-4351-8eb1-8126185aeea2-0 So they just push things to Kafka and then they expect it to 76476435-e1c0-4351-8eb1-8126185aeea2-1 be the source of truth. c3d5e405-2e4e-42c2-aa4c-3a84a3b8eaad-0 They expect it to be the place where they can fetch all the c3d5e405-2e4e-42c2-aa4c-3a84a3b8eaad-1 reporting data that can do all of their order processing and c3d5e405-2e4e-42c2-aa4c-3a84a3b8eaad-2 they can do all other metrics. 44a20452-593c-4f80-bb06-567b03e81a06-0 Sometimes they might add a little layer in between, but 44a20452-593c-4f80-bb06-567b03e81a06-1 this is what we've seen before and this is not the solution. 466c2d9d-b38a-467b-b0ee-5252baab5342-0 Kafka is not a database and it is not the way to solve 466c2d9d-b38a-467b-b0ee-5252baab5342-1 that particular problem or really any problem from a 466c2d9d-b38a-467b-b0ee-5252baab5342-2 accessing Kafka directly tohelp with your scalability and 466c2d9d-b38a-467b-b0ee-5252baab5342-3 speed issues. b0b3e845-a526-4b46-9108-7af1e6557e41-0 So before we dive into you know how to solve it, I want to talk b0b3e845-a526-4b46-9108-7af1e6557e41-1 a little bit about databases. d1c30290-3756-4d49-a221-42ded2caabc2-0 So traditionally when you hear database, you think of SQL or no d1c30290-3756-4d49-a221-42ded2caabc2-1 SQL databases, you think of a database that has a secondary d1c30290-3756-4d49-a221-42ded2caabc2-2 index. df2cd1ba-419d-40f0-9db6-d67a244a7be8-0 Databases have tables and one of themost traditional ways to df2cd1ba-419d-40f0-9db6-d67a244a7be8-1 think about a database is that it's ACID compliant. ecc9f14a-86cb-4db3-b15f-40e28282a9de-0 So it's atomicity, consistency, isolation and durability. 78b56e98-9491-4770-9be8-9c9f5f459c2d-0 And that really applies to relational databases. 2d00e0a8-14ea-41db-ad2c-368f9cb2b5e8-0 And in this list of technologies that are often called databases 2d00e0a8-14ea-41db-ad2c-368f9cb2b5e8-1 actually Rocks database and Redis, they're not ACID 2d00e0a8-14ea-41db-ad2c-368f9cb2b5e8-2 compliant, but you can see Rocks DB has DB right there in the 2d00e0a8-14ea-41db-ad2c-368f9cb2b5e8-3 name, right? ae406e19-3de0-47d6-9174-c529e05f4a28-0 And Redis is Redis cash, but it's often referenced as a ae406e19-3de0-47d6-9174-c529e05f4a28-1 database. 59c9b1f2-7853-4ffc-bea9-5ebf8110c6d8-0 Even Cassandra is only eventually consistent. 85d7a1b3-8e4a-4243-89f7-b5e3d638c886-0 So it is a database, but the term isgetting a little more 85d7a1b3-8e4a-4243-89f7-b5e3d638c886-1 loosely defined andthat you know it can be in challenge when 85d7a1b3-8e4a-4243-89f7-b5e3d638c886-2 you talk about Apache Kafka as a database. 39426418-525e-44f7-b265-419757ad73b6-0 So I've said it once already or a couple times, is Kafka a 39426418-525e-44f7-b265-419757ad73b6-1 database? cf8af840-dd89-4e05-9e90-67b2daf04c44-0 I,personally don't believe it is. ef12c6f1-23a1-4334-bca2-3f6193746872-0 And I think the community at large also doesn't believe that ef12c6f1-23a1-4334-bca2-3f6193746872-1 it in itself is a database. 3f2001d8-13a0-4b76-8e2a-7c8fbdce46cf-0 It doesn't have a query language that you can access when you're, 3f2001d8-13a0-4b76-8e2a-7c8fbdce46cf-1 when you're accessing data from Apache Kafka, you're3f2001d8-13a0-4b76-8e2a-7c8fbdce46cf-2 accessing a write ahead log, right? 52bdb0ef-5ba0-4973-82d4-eb59d1eb1448-0 So, you know, you can definitely request a specific, you know, 52bdb0ef-5ba0-4973-82d4-eb59d1eb1448-1 records from a specific time frame. 64796136-8747-428f-a10a-baf5348fa96d-0 You can request records based on, you know, a start and stop 64796136-8747-428f-a10a-baf5348fa96d-1 offset when you key your data. 2933689d-686e-4253-af8f-49cfaaeec673-0 You can, you know, request certain keys. 15e62fe5-838e-4dab-a087-1d9ecc4af7cf-0 But if you consider it, you know, as a, as a key value pair, 15e62fe5-838e-4dab-a087-1d9ecc4af7cf-1 type database or a type cash, it just isn't there, right? a22a10be-0d06-4b91-8b8d-079bdfda59bf-0 It isn't there when it comes to speed for the most part. 6006956b-c0df-4c32-a7d1-a24c85e75f71-0 So there's no, there's no indices in Apache Kafka. 65ffcaef-2cd1-4446-88b0-43364b669bb0-0 It's just a key. def53faa-0d49-4406-a7ce-194ff7c7ef5a-0 I mean, I guess you could argue that the offset servers and def53faa-0d49-4406-a7ce-194ff7c7ef5a-1 index, but where would you track with which offset stores what, def53faa-0d49-4406-a7ce-194ff7c7ef5a-2 right? 1d285663-152f-4407-b759-cbf649517b7c-0 Andthat wouldn't be in Kafka specifically, which even 1d285663-152f-4407-b759-cbf649517b7c-1 lends more weight to that ititself is not a database. 2dacc232-9614-4368-b8f7-3517d057c811-0 A topic is not the same thing as a table, right? d9da69e2-a4a3-4216-b507-42d67cae2dd2-0 Sure, it's kind of a key value pair with some added metadata, d9da69e2-a4a3-4216-b507-42d67cae2dd2-1 but it is not a table. 6c5f5559-2cd6-4286-9971-2005b703e8d8-0 You're not going to apply secondary indexes to a Kafka 6c5f5559-2cd6-4286-9971-2005b703e8d8-1 topic. b4257b9e-10bc-487a-b904-6450a0ef1199-0 It's just going to be thatprimary,key. bc7fe886-0042-4626-b134-0ff323a33e54-0 And lastly, Kafka is not ACID compliant. 3ca97e98-0d32-4352-8634-361b4ce73a77-0 And you know, I think I could go through the explanation here, 3ca97e98-0d32-4352-8634-361b4ce73a77-1 but honestly, Martin Klupman in the Kafka Summit of 2019, he did 3ca97e98-0d32-4352-8634-361b4ce73a77-2 a really good job of kind of going through each,one of 3ca97e98-0d32-4352-8634-361b4ce73a77-3 those and,how Kafka and other technology solves those 3ca97e98-0d32-4352-8634-361b4ce73a77-4 problems. 2db1571c-1d88-4cf1-aa5a-4a615e11f33d-0 And in the end, his conclusion was the same, which is that 2db1571c-1d88-4cf1-aa5a-4a615e11f33d-1 Kafka is not a database, but you could use Kafka to kind of 2db1571c-1d88-4cf1-aa5a-4a615e11f33d-2 create your own version of a database if you needed to, 2db1571c-1d88-4cf1-aa5a-4a615e11f33d-3 right? 121fbff3-f7cb-44cb-81a1-02b5a6d54c17-0 And that was kind of his in conclusion, which I agree with. 3a5696f5-6392-455c-a453-629d7b806ccd-0 But in the end, is the juice worth the squeeze, right? 380abd75-825e-4b80-bdd6-429b882e4b9d-0 Do you have the need for that? 6e68f36d-a815-4db4-bbc9-59d7c5cbbfde-0 Do you have the expertise to understand that and to absorb 6e68f36d-a815-4db4-bbc9-59d7c5cbbfde-1 that technical debt long term? 5bb3aa16-bba3-4da7-bc58-d8e9f2ccd98c-0 Because while you might be able to make something today that 5bb3aa16-bba3-4da7-bc58-d8e9f2ccd98c-1 that can act like a database, is that maintainable long term for 5bb3aa16-bba3-4da7-bc58-d8e9f2ccd98c-2 you? ad652025-4f9b-4997-9ab8-84ff0385ee2f-0 And for most people, the answer to that is no. fc02cb1d-324c-40d0-96a8-85682c0f744b-0 So what is Kafka? aa94db36-6c63-4d33-bf8c-09e15d308dd0-0 Well, Kafka is a distributed event streaming platform, right? 49c56f77-7bd4-4b4f-80a4-dbecb8284bae-0 That was the mission that it set out to get. eba2be2c-9186-486d-8c00-9c627d54cff2-0 It says have and that's what it is today. f0fef745-a5d8-4e3c-a911-bf7b0777b66d-0 It can be used as a data store, but it does have trade-offs, f0fef745-a5d8-4e3c-a911-bf7b0777b66d-1 right? de6fc250-81b5-468b-9ff5-52e3c9b7425d-0 You can keep data in there for a long time, have lots of de6fc250-81b5-468b-9ff5-52e3c9b7425d-1 retention, but the way you access that data, you know, de6fc250-81b5-468b-9ff5-52e3c9b7425d-2 there's,trade-offs. bfa03326-717f-4f2c-aa7a-d906b43949ea-0 And that's important to understand when you,know, bfa03326-717f-4f2c-aa7a-d906b43949ea-1 when you choose your technology for the solution that you're bfa03326-717f-4f2c-aa7a-d906b43949ea-2 trying to solve or for the problem that you're trying to bfa03326-717f-4f2c-aa7a-d906b43949ea-3 solve. dca17cdc-7094-423e-be6e-ab16c8cd1be7-0 It does have some features consistent with other dca17cdc-7094-423e-be6e-ab16c8cd1be7-1 distributed technologies or databases, but itself by itself dca17cdc-7094-423e-be6e-ab16c8cd1be7-2 is not a database. 2e0cf0a6-29d1-48df-aa06-bc723ac49a0f-0 You could use tools in the Kafka ecosystem, one of those being 2e0cf0a6-29d1-48df-aa06-bc723ac49a0f-1 like K SQL DB to make it act a little more like a database. 8b35b295-41c8-4124-832f-9729e27691a5-0 But from my experience and I have some friends in the 8b35b295-41c8-4124-832f-9729e27691a5-1 industry whohave tried to use K SQL DB and it works in 8b35b295-41c8-4124-832f-9729e27691a5-2 small to medium sized scales. a4c1dce8-f8d3-4916-8495-1ce325afaff9-0 But once you really scale up and you have high velocity data a4c1dce8-f8d3-4916-8495-1ce325afaff9-1 though it just it can't keep up withbeing performant and a4c1dce8-f8d3-4916-8495-1ce325afaff9-2 in answering your requests in a what most people would a4c1dce8-f8d3-4916-8495-1ce325afaff9-3 consider a reasonable amount of time once you reach a high a4c1dce8-f8d3-4916-8495-1ce325afaff9-4 velocity of data and a very high volume of data. 65753ebd-8c40-4029-a8c7-3ed21c508b39-0 But for those very small low velocity use cases or,even 65753ebd-8c40-4029-a8c7-3ed21c508b39-1 some medium cases, it can work, you know, butnot for the 65753ebd-8c40-4029-a8c7-3ed21c508b39-2 high velocity. 4960afaa-02dc-456d-aac2-054c515bb4e8-0 And most people who are using Apache Kafka, they're,4960afaa-02dc-456d-aac2-054c515bb4e8-1 in that high velocity phase, right? 202e53e9-059b-4a07-9c75-b20820a266c4-0 Andthat's where they really have a need to do those look 202e53e9-059b-4a07-9c75-b20820a266c4-1 UPS. 6a49bb2b-6943-4df0-aa5a-f2b312939231-0 So back to our oversimplified use case on,selling some 6a49bb2b-6943-4df0-aa5a-f2b312939231-1 stuff. 168499ac-8632-4cb5-bad4-8770566af710-0 How can I use Apache Kafka, right? 7b3bde0b-d19f-428f-9b57-47cbc2b080b7-0 So where do we go? 665f137e-bfeb-411a-a642-e277a0f804e4-0 Well, so again, you've got your users and a website and a 665f137e-bfeb-411a-a642-e277a0f804e4-1 website hits an API. 45bd9e57-61ab-4fc6-9c73-0d34ef7c1965-0 So instead of that API saving everything to a traditional 45bd9e57-61ab-4fc6-9c73-0d34ef7c1965-1 relational database, we can envision that being pushed to 45bd9e57-61ab-4fc6-9c73-0d34ef7c1965-2 Apache Kafka as events. ccc9f6ee-2fd3-4e02-a843-c08b90db780a-0 And so again, event streaming is the concept here. 44ebf977-05e7-47a8-9958-d325a2708bda-0 And so imagine in this, in Kafka, now you have all of your 44ebf977-05e7-47a8-9958-d325a2708bda-1 events, whether that's customer saving their profiles, placing 44ebf977-05e7-47a8-9958-d325a2708bda-2 orders, any sort of click stream data, right? aef08db8-e2ea-4e9d-b236-398072b1e24d-0 It's all segmented in Apache Kafka so that you can at on each aef08db8-e2ea-4e9d-b236-398072b1e24d-1 piece of data in near real time andeach,component of aef08db8-e2ea-4e9d-b236-398072b1e24d-2 that is independent, right? b5fed184-049a-484d-b2fa-8bc0bff940df-0 So it functions on its own. 028498ba-a75f-4c12-9f94-2bd0f2a3080c-0 So in this use case, you,can have one or more jobs. cfbe652a-bd31-497b-809a-f59fc4d5bdd0-0 You can have, you could have two jobs here, right? 34353f3a-57a5-4ad8-aa7b-4426dd7342c6-0 You have one that saves customer data and one that saves order 34353f3a-57a5-4ad8-aa7b-4426dd7342c6-1 data, right? 7d6fc4f8-b397-4381-948b-8dd6088c4cfb-0 And by save, we mean, let's put it in a database, right? 647315ff-4e14-4699-b27b-8f92349a41bf-0 There's no reason to go away from a database. 888a733c-0f4b-493b-aa1f-b6c7fa5ce0e0-0 Apache Kafka is here to helpyou improve your 888a733c-0f4b-493b-aa1f-b6c7fa5ce0e0-1 data-driven ecosystem. 87c40e15-1b98-4cd2-ab76-c6c20dbf7aab-0 And you can still use the tools that you know and that you're 87c40e15-1b98-4cd2-ab76-c6c20dbf7aab-1 familiar with because relational databases are everywhere. 45c761f6-c732-41da-8bac-bd30b047c081-0 They're ubiquitous and,being able to save data into a 45c761f6-c732-41da-8bac-bd30b047c081-1 relational database if it's not super high velocity andor 45c761f6-c732-41da-8bac-bd30b047c081-2 very large volumes still has a lot of value because so many 45c761f6-c732-41da-8bac-bd30b047c081-3 people know how to use it and it's so easy to use. 18859301-0d7c-4435-bab5-0b08e9eec511-0 So there's nothing wrong with using that database. 6c32ebba-402a-437a-a7d9-115a8713b6e7-0 So drop the drop that type of data in your database and then 6c32ebba-402a-437a-a7d9-115a8713b6e7-1 use it to write some reports, right. da92ec9f-9f5b-4309-af6c-83c4a351dd06-0 If you need still need those daily, weekly roll ups, you can da92ec9f-9f5b-4309-af6c-83c4a351dd06-1 still get your customers and your order data and feed it from da92ec9f-9f5b-4309-af6c-83c4a351dd06-2 there. ea2346e7-c6b0-4505-8823-33e32446edf9-0 You can even drive CRM systems or other various systems based ea2346e7-c6b0-4505-8823-33e32446edf9-1 on that relational database. 5bb75f7e-ff83-4555-aa12-bf5784aec0a0-0 There's nothing wrong with that. 9a401858-7c5a-4ce9-858e-0837facb6d29-0 Where Kafka really helps is when those new use cases come up that 9a401858-7c5a-4ce9-858e-0837facb6d29-1 we talked about a few slides back. 63f9d7f2-e618-427b-883f-45d3bab34d25-0 Now all of a sudden you need to do some data aggregation. bab3e76b-0a0e-4a70-afc4-ce0ba5f51122-0 Well, you can use a tool in the Kafka ecosystem, maybe it's AK bab3e76b-0a0e-4a70-afc4-ce0ba5f51122-1 Streams Job or Kafka Connect and you can do data aggregations. 3f4e1fcf-3296-40a9-8e59-e7aa9a6cf3fe-0 And those data aggregations, they may live in a 3f4e1fcf-3296-40a9-8e59-e7aa9a6cf3fe-1 relational database, or you may be able to just aggregate data 3f4e1fcf-3296-40a9-8e59-e7aa9a6cf3fe-2 in, you know, in a time series or a windowed fashion and push 3f4e1fcf-3296-40a9-8e59-e7aa9a6cf3fe-3 that to a reporting engine that provides your metrics like a 3f4e1fcf-3296-40a9-8e59-e7aa9a6cf3fe-4 Tableau or click or something like that. 1e953d0d-c1ba-4c52-987d-e5a453e605ef-0 And so you're aggregating data and you're pushing it out to 1e953d0d-c1ba-4c52-987d-e5a453e605ef-1 something to provide you real time or near real time metrics, 1e953d0d-c1ba-4c52-987d-e5a453e605ef-2 right? dec070a8-4865-430f-83bb-d593365d9ce2-0 Well, what if I want to save some data to cache to maybe dec070a8-4865-430f-83bb-d593365d9ce2-1 power ACRM tool or a website or something like that? 6fd19e57-ccd4-46ea-a24b-2a9822d08cba-0 Again, you can just plug in that new feature right into Apache 6fd19e57-ccd4-46ea-a24b-2a9822d08cba-1 Kafka because all of your events live there and you can create a 6fd19e57-ccd4-46ea-a24b-2a9822d08cba-2 job that saves it to cache. 55ed6a6c-b962-47d3-9ebc-2d5c79bbf6a5-0 Maybe your cache is Redis. b4770d81-233a-4065-a7aa-b4851de8b1ef-0 Maybe your cache is a compacted topic in Kafka. 2774be36-56ea-4a63-aebc-ff5c65b6f015-0 That's going to be a decision that you have to make based on 2774be36-56ea-4a63-aebc-ff5c65b6f015-1 your data access patterns andyour needs of that cash. a2e4e6a9-02f6-4e72-911b-95377fb01acf-0 And again, we talk about the plug ability of Apache Kafka. 1c6dbef2-d135-45ce-8614-08faa09d8755-0 And so you can just plug in a lot of various components into 1c6dbef2-d135-45ce-8614-08faa09d8755-1 Kafka and you're not going to see degradation and service 1c6dbef2-d135-45ce-8614-08faa09d8755-2 because it's super scalable, it's reliable, it's highly 1c6dbef2-d135-45ce-8614-08faa09d8755-3 available. 079bb410-be5f-4a51-a3a3-c9f7e93c4d7c-0 And,so you can use it as the heart of your data 079bb410-be5f-4a51-a3a3-c9f7e93c4d7c-1 ecosystem, but not as a database. 325eab1e-8cd7-4202-a805-32539a30e868-0 It is complementary of your databases. 775ef8ee-ff3f-4ac9-b4b3-2c7cdcc8a44e-0 And I don't have it here, but you can put data lakes in here. 7ed7eb29-b401-4b3e-9438-0543726dbe13-0 You can, you can feed these data to, you know, what, whatever 7ed7eb29-b401-4b3e-9438-0543726dbe13-1 type of lake or,big data platform that you want to want 7ed7eb29-b401-4b3e-9438-0543726dbe13-2 to keep it in Apache Cassandra or something like that. 502ce29b-c5c5-47a8-8473-28734aac2804-0 Andthen your, you know, your AI guys, your e-mail guys, 502ce29b-c5c5-47a8-8473-28734aac2804-1 your people working on your,big data stuff. 9addf12b-6d30-4876-bf46-57f322f1a438-0 They can pull that data fromeither Kafka directly, if 9addf12b-6d30-4876-bf46-57f322f1a438-1 they have real time jobs that are running or if they have jobs 9addf12b-6d30-4876-bf46-57f322f1a438-2 that need to run on historical data over a week or a month, you 9addf12b-6d30-4876-bf46-57f322f1a438-3 can just pipe that data into whatever, you know, place that 9addf12b-6d30-4876-bf46-57f322f1a438-4 you want it. c606b903-4795-4226-a5eb-f7914c18f3c1-0 And again, your Kafka is not going to take a big impact from c606b903-4795-4226-a5eb-f7914c18f3c1-1 that because it serves as that message broker. f9eb6e3a-c980-4339-bf18-c09a116b6005-0 It is that central pipe for,your ecosystem of data. a5498450-84eb-4b83-a71b-1e954b08ddb1-0 And that's all I had. 21e968aa-0d85-4fd5-be3c-e1e9d9166380-0 And so we'll,leave it open to some Q&A. 1e64dfa1-bc98-44d4-9bed-cd47f46bc18c-0 I know it was a short presentation, but I'm sure 1e64dfa1-bc98-44d4-9bed-cd47f46bc18c-1 Stefan's got a,few questions in there. 3ddfc6db-bcde-4756-bc7d-740af43dfd1a-0 Thanks, Andrew for the presentation. 36a5b3bc-3144-4422-8daa-64dac98b71b7-0 See, like Andrew said, we'll be jumping into just some questions 36a5b3bc-3144-4422-8daa-64dac98b71b7-1 that we received from the audience. 6cc0a147-e749-4a7e-9e49-408339d65ae3-0 If you haven't had a question already for you to do so now, 6cc0a147-e749-4a7e-9e49-408339d65ae3-1 and we'll try to eat as many as we can. 5f98ecb1-4350-4441-ba2a-d4fb9abd5354-0 Let's see. f7f5b906-7cfe-4bb0-a0af-301a451188fa-0 So one question I see here is, Andrew, if you use Kafka as a f7f5b906-7cfe-4bb0-a0af-301a451188fa-1 cache, is that a database? 9a5de8f9-c8c8-4939-963b-1456d1b728d0-0 Back to the definition of a database right. 4261c6e7-6103-4f2c-ab8a-c609aff9d324-0 So ifby database you mean ACID compliant no. 80382158-7532-420f-b399-aac69d664b39-0 But if you think about you know like rocks DB, which is more a 80382158-7532-420f-b399-aac69d664b39-1 lot of the times like a cache or Redis cache, right. 09874722-669b-4323-bb75-f7ee31ac2496-0 If you consider that to be database, you know Iwouldn't 09874722-669b-4323-bb75-f7ee31ac2496-1 consider that itbeing a database but. 95cc757d-564c-49cc-8e5b-a097c1fd56d2-0 Compacted topics in Kafka, I would say act more like a 95cc757d-564c-49cc-8e5b-a097c1fd56d2-1 database thanKafka itself, because youroll up the data 95cc757d-564c-49cc-8e5b-a097c1fd56d2-2 into a key. 7b8f0c02-140f-497b-b62d-5bf3d3338e6b-0 Roll up isn't the right word, sorry. 552b858e-0acf-4529-8e0c-a559116e7392-0 You,get to replace the old data, rewrite the data with552b858e-0acf-4529-8e0c-a559116e7392-1 some with a new event based on the key. 847db7a2-e20e-47a7-b719-b907bc7ca515-0 And it's not, you're not going to do a traditional CRUD type 847db7a2-e20e-47a7-b719-b907bc7ca515-1 operation. c398e73c-3295-4d7d-b7c3-5fe9faba3217-0 It's just a straight replace because whatever,you c398e73c-3295-4d7d-b7c3-5fe9faba3217-1 write to that key next is what overwrites it. c219137d-c63e-45bb-a2ba-fef2bac2528b-0 And so does it act kind of like a database? d1a521ce-cbb8-466b-b5e1-05add4c22e77-0 Sure. 79279024-7712-4944-bc33-29692638b438-0 But it's,not the same thing. 98261876-25a7-41e2-b765-a85f0fabe8f0-0 And I think what the most important thing to consider is 98261876-25a7-41e2-b765-a85f0fabe8f0-1 you have to understand your data access patterns and the 98261876-25a7-41e2-b765-a85f0fabe8f0-2 trade-offs, right? eb18e825-0010-4a36-b496-bb013be7854a-0 So if a compacted topic is right for you, for your use case, then eb18e825-0010-4a36-b496-bb013be7854a-1 then use it. 0347ac6b-37d9-419e-b3d9-62586051881c-0 There's no reason to pull another technology if it'll 0347ac6b-37d9-419e-b3d9-62586051881c-1 work. 8547c3da-9e3f-4dde-ae95-35f221e01b7f-0 But if you're expecting a compacted topic to serve you 8547c3da-9e3f-4dde-ae95-35f221e01b7f-1 records as fast as a Redis cash will, you're probably going to 8547c3da-9e3f-4dde-ae95-35f221e01b7f-2 fall short on those expectations. 5445d10d-31c2-43fb-9adc-66daf7f57c07-0 I hope that answered your question. 906bbb48-3165-49af-87f5-26844a9a10c7-0 Thanks Andrew for answering that one. ad3c59ae-5887-41c7-acfa-bc77df89a88f-0 See another question I see here is someone asks what if I'm ad3c59ae-5887-41c7-acfa-bc77df89a88f-1 using Kafka as a database Now can I go back? 995986dc-32b4-40ff-81c2-deee4d528752-0 I mean, sure, I think the same principle applies, 995986dc-32b4-40ff-81c2-deee4d528752-1 which is understanding your access patterns andis it 995986dc-32b4-40ff-81c2-deee4d528752-2 working for you. 41ba4117-d540-4992-a19b-9fcd3adf7957-0 So look, if,you, if you did, you know the use case that we 41ba4117-d540-4992-a19b-9fcd3adf7957-1 said andyou've put Kafka at the heart and everything's 41ba4117-d540-4992-a19b-9fcd3adf7957-2 working fine. 3f5cf072-763d-4f67-8f8d-afbf5a5fceab-0 There's no reason to make change for change sake, right? 88f2b471-3b54-4810-a418-1cd2dddf9e97-0 If it ain't broke, don't fix it, right? f4c5486d-27b1-4b88-b1fa-83e49529ffd5-0 But if you intend to scale, you intend for your business to grow f4c5486d-27b1-4b88-b1fa-83e49529ffd5-1 and you're you expect more traffic, then I would say f4c5486d-27b1-4b88-b1fa-83e49529ffd5-2 consider changing the way that you're using it to be more in f4c5486d-27b1-4b88-b1fa-83e49529ffd5-3 line with the,kind of the last pattern that I showed. 01e2f62a-19b7-4b94-91da-6fbb8ab8dfed-0 Andinstead of using Kafka itself as your database, begin 01e2f62a-19b7-4b94-91da-6fbb8ab8dfed-1 to use other systems at their strength, right? ea0e48ec-3432-4bcd-b1d6-ee7f3997093c-0 So if relational database, for instance, is,really strong ea0e48ec-3432-4bcd-b1d6-ee7f3997093c-1 in certain ways and,start moving that direction because ea0e48ec-3432-4bcd-b1d6-ee7f3997093c-2 once you do scale it, it's really hard to start solving ea0e48ec-3432-4bcd-b1d6-ee7f3997093c-3 problems once you start to see them. b10bb39e-f133-4f3e-ad2e-11cb97db207a-0 I think the most common practice is throw more hardware at it b10bb39e-f133-4f3e-ad2e-11cb97db207a-1 instead of diving into the architecture of either you know, b10bb39e-f133-4f3e-ad2e-11cb97db207a-2 the system or how it uses, you know, JVMS or how it uses the, b10bb39e-f133-4f3e-ad2e-11cb97db207a-3 you know, Linux or understanding deeply, you know how your b10bb39e-f133-4f3e-ad2e-11cb97db207a-4 partitioning is working in Kafka. e66fbff8-02f6-4a98-b6f0-b5031df0bfe9-0 So I,would say, yes, you can go back. 510f1d4d-c67f-4f8f-9ab7-3b366cd5ea06-0 I don't know if that's the right term, but youcan definitely 510f1d4d-c67f-4f8f-9ab7-3b366cd5ea06-1 move away from using it as a database and byimplementing, 510f1d4d-c67f-4f8f-9ab7-3b366cd5ea06-2 you know, other technologies where they're the strongest and 510f1d4d-c67f-4f8f-9ab7-3b366cd5ea06-3 and simply usingKafka as you know, the heart of your 510f1d4d-c67f-4f8f-9ab7-3b366cd5ea06-4 system. 13ef0eef-400e-4805-8850-5b626b10a43e-0 So you definitely made a good choice in putting Kafka at the 13ef0eef-400e-4805-8850-5b626b10a43e-1 heart, right? b9344b92-b368-4b34-af59-b4828c981568-0 But you just have to branch out a little bit on your other b9344b92-b368-4b34-af59-b4828c981568-1 technologies. 67b2beba-ab68-4138-b95e-6620acd56cc9-0 Hope that answer your question. b36409bb-7bd0-4ae4-a7b0-4b5cb5561535-0 All right, looks like we've time for just one more question this b36409bb-7bd0-4ae4-a7b0-4b5cb5561535-1 one something maybe touched a little bit on during post b36409bb-7bd0-4ae4-a7b0-4b5cb5561535-2 station. 2047c28e-6b19-465c-9dea-4558bd5ed922-0 Andrew, this person asks what kind of scaling issues have you 2047c28e-6b19-465c-9dea-4558bd5ed922-1 seen with K sequel DB? 721a2391-94ca-4591-bb19-9a0ebf6de8cb-0 Yeah. 408f5b3d-fa6a-445d-8f31-d6235f4cbd43-0 So, you know, I've used it sparingly, I would say 408f5b3d-fa6a-445d-8f31-d6235f4cbd43-1 definitely more than hello world, butI've used it in 408f5b3d-fa6a-445d-8f31-d6235f4cbd43-2 that small to medium sized pattern and you know, it, it's 408f5b3d-fa6a-445d-8f31-d6235f4cbd43-3 not bad. 64cdd687-4ee1-46c4-9a7f-dc04a8c627b1-0 I mean, essentially the layer that it sits thatit 64cdd687-4ee1-46c4-9a7f-dc04a8c627b1-1 provides, sits on top of Kafka. d77d380f-ed95-4b1d-a8da-198b166b13e1-0 It,does, it does provide value, but I have some friends d77d380f-ed95-4b1d-a8da-198b166b13e1-1 that work at a large fintech and they,tried to deploy it in d77d380f-ed95-4b1d-a8da-198b166b13e1-2 a in a very large capacity and it just, it didn't, it didn't d77d380f-ed95-4b1d-a8da-198b166b13e1-3 function like, you know, you weren't getting any results back d77d380f-ed95-4b1d-a8da-198b166b13e1-4 on your queries. f69306ee-3c61-40bf-809f-7f2dc09ba7bf-0 Not,just in a timely fashion at all, right. cf7c894d-6754-4612-9577-775a928a66fa-0 They were getting, you know, overflow errors and,they cf7c894d-6754-4612-9577-775a928a66fa-1 worked with some,consultants that are very well cf7c894d-6754-4612-9577-775a928a66fa-2 known in the industry to,try to understand, you know, what cf7c894d-6754-4612-9577-775a928a66fa-3 was happening. 682b8484-c336-4b2b-9479-2c7dda248acf-0 And in the end, after weeks and weeks of troubleshooting, K 682b8484-c336-4b2b-9479-2c7dda248acf-1 Sequel DB just didn't work for that use case, despite various, 682b8484-c336-4b2b-9479-2c7dda248acf-2 you know, changes in patterns and the way they were accessing 682b8484-c336-4b2b-9479-2c7dda248acf-3 things and,different use case testing and whatnot. 902a07a3-fa1e-4758-a91b-19dca8e86bde-0 It ata high level scale with,high velocity of 902a07a3-fa1e-4758-a91b-19dca8e86bde-1 data. a0182f4e-8475-417c-a38c-20495b46cade-0 I think that's probably the most important part is the high a0182f4e-8475-417c-a38c-20495b46cade-1 velocity of data. e6c62198-cc77-4d70-8a13-a1159ffdc202-0 It,just is not a viable solution to,be able to e6c62198-cc77-4d70-8a13-a1159ffdc202-1 quickly access your data like you would a relational database. 39b2d09c-77fa-4e3d-b18c-a8e00ef048ec-0 So thatwould be the extent I would say on that since it's 39b2d09c-77fa-4e3d-b18c-a8e00ef048ec-1 not my personal experience, but a good friend of mine who had 39b2d09c-77fa-4e3d-b18c-a8e00ef048ec-2 that experience. d347329a-47a9-4853-bbc6-3de21f79759e-0 We talked at length about it. 579920c4-98fc-43bc-96dc-9e9418e4aaa6-0 So all right, so it's like that with that last question. 4059d5bf-a65a-4f8b-ac58-7f20ddd86cf1-0 That concludes our presentation. 47e25542-0daa-4c28-9f59-8e1148cbd5c6-0 I just want to thank everyone who types in their questions. 696d60a5-e297-4fdf-b20d-de9917ef19f6-0 Sorry if we couldn't get to all of them today, but we will be 696d60a5-e297-4fdf-b20d-de9917ef19f6-1 reaching out directly via e-mail to ensure that yours are 696d60a5-e297-4fdf-b20d-de9917ef19f6-2 answered. 109616e8-eaef-4217-ada2-611815ae5aff-0 In the meantime, if you do have any other questions that pop up, 109616e8-eaef-4217-ada2-611815ae5aff-1 you can e-mail us at info@inscluster.com and we will 109616e8-eaef-4217-ada2-611815ae5aff-2 be happy to get back to you that way as well. d5d93018-7829-4721-8aac-eba7acabf53a-0 In addition, if you're interested in learning more d5d93018-7829-4721-8aac-eba7acabf53a-1 about Instacluster and the work we do with Apache Kafka, you can d5d93018-7829-4721-8aac-eba7acabf53a-2 visit our website instacluster.com where you have d5d93018-7829-4721-8aac-eba7acabf53a-3 numerous blogs and help resources available around d5d93018-7829-4721-8aac-eba7acabf53a-4 Kafka. 5a014aa5-9714-4aa7-b143-14b61f945499-0 So I definitely recommend checking that out. 7aeb3c5b-aa2d-4717-8a8f-6e20b880119c-0 So as a wrap up, I just want to thank Andrew again for his time 7aeb3c5b-aa2d-4717-8a8f-6e20b880119c-1 today and thanks for all the attendees for listening. 5b63f6b4-4ea9-4477-b15b-16c1e06667d0-0 Good to see you in our next webinar. 254e6304-8d60-4d04-8fee-3040188a32f2-0 Thanks and have a great day. c7d81b2c-573f-479c-aa11-753008adeb1a-0 Thanks folks.