WEBVTT 1 00:00:00.180 --> 00:00:01.160 Let's talk about what I think 2 00:00:01.160 --> 00:00:02.570 is one of the most exciting parts 3 00:00:02.570 --> 00:00:04.650 of the Redis modules, RedisJSON. 4 00:00:04.650 --> 00:00:06.830 This basically turns Redis into a full-featured, 5 00:00:06.830 --> 00:00:10.670 horizontally-scalable NoSQL document store. 6 00:00:10.670 --> 00:00:11.900 That's a big deal, right? 7 00:00:11.900 --> 00:00:14.800 I mean, instead of using some external third-party system 8 00:00:14.800 --> 00:00:15.840 and its own complexity 9 00:00:15.840 --> 00:00:18.540 and having to understand how it works and is configured, 10 00:00:18.540 --> 00:00:21.060 you can do all this within Redis now. 11 00:00:21.060 --> 00:00:22.790 So the nice thing about RedisJSON 12 00:00:22.790 --> 00:00:25.330 is that it allows you to update individual fields 13 00:00:25.330 --> 00:00:27.730 within your structured data atomically, right? 14 00:00:27.730 --> 00:00:29.157 So you might say to yourself, 15 00:00:29.157 --> 00:00:31.060 "Hey, what's the big deal about storing JSON?" 16 00:00:31.060 --> 00:00:33.430 It's just a string at the end of the day, right? 17 00:00:33.430 --> 00:00:34.650 But think about that, 18 00:00:34.650 --> 00:00:36.940 like if you really just wanted to update one field 19 00:00:36.940 --> 00:00:38.350 in a JSON structure, 20 00:00:38.350 --> 00:00:39.360 without RedisJSON, 21 00:00:39.360 --> 00:00:42.300 you'd have to read back that entire JSON structure, 22 00:00:42.300 --> 00:00:43.310 change the piece that you want, 23 00:00:43.310 --> 00:00:44.610 and write the whole thing back. 24 00:00:44.610 --> 00:00:46.990 And parsing that could be time consuming 25 00:00:46.990 --> 00:00:48.470 and resource intensive. 26 00:00:48.470 --> 00:00:50.690 With RedisJSON, you don't have to deal with that. 27 00:00:50.690 --> 00:00:51.523 You can just say, 28 00:00:51.523 --> 00:00:54.270 "Hey, I want to update this piece of my data structure, 29 00:00:54.270 --> 00:00:55.680 go deal with it, RedisJSON, 30 00:00:55.680 --> 00:00:58.500 and do it atomically while you're at it." 31 00:00:58.500 --> 00:00:59.950 Let's look at some examples. 32 00:00:59.950 --> 00:01:03.460 So here's an example JSON data structure, very simple. 33 00:01:03.460 --> 00:01:05.410 Let's say that I'm keeping a little database 34 00:01:05.410 --> 00:01:07.610 of "Star Trek" ships, 35 00:01:07.610 --> 00:01:08.570 and we want to keep track 36 00:01:08.570 --> 00:01:11.750 of what the captains and designations of those ships are. 37 00:01:11.750 --> 00:01:13.540 So for the ship enterprise, 38 00:01:13.540 --> 00:01:18.540 the captain is kirk and its designation is NCC-1701. 39 00:01:18.560 --> 00:01:19.393 How would you do that? 40 00:01:19.393 --> 00:01:22.030 So we could just say, JSON.SET, 41 00:01:22.030 --> 00:01:24.570 these are actual Redis CLI commands here, 42 00:01:24.570 --> 00:01:27.490 enterprise.'"captain" : "kirk"' 43 00:01:27.490 --> 00:01:29.540 So that sets things up with our initial data there 44 00:01:29.540 --> 00:01:32.530 of we're mapping captain and kirk 45 00:01:32.530 --> 00:01:33.850 under the structure enterprise. 46 00:01:33.850 --> 00:01:35.880 So already we have this nested data structure there 47 00:01:35.880 --> 00:01:37.470 with one line of code. 48 00:01:37.470 --> 00:01:39.060 I can get that data back very quickly 49 00:01:39.060 --> 00:01:41.910 using just JSON.GET enterprise. 50 00:01:41.910 --> 00:01:43.900 Again, just captain kirk is what comes back. 51 00:01:43.900 --> 00:01:45.340 That's what you'd expect. 52 00:01:45.340 --> 00:01:47.490 Now let's say I want other add some new data 53 00:01:47.490 --> 00:01:48.510 to that structure. 54 00:01:48.510 --> 00:01:50.040 I can do that with one line as well, 55 00:01:50.040 --> 00:01:53.090 and again, without having to retrieve the whole thing first. 56 00:01:53.090 --> 00:01:57.470 So I could just say JSON.SET enterprise .designation 57 00:01:57.470 --> 00:01:59.300 to add that designation field, 58 00:01:59.300 --> 00:02:01.840 and I'll give it a value of "NCC-1701". 59 00:02:01.840 --> 00:02:03.860 Pretty simple syntax there. 60 00:02:03.860 --> 00:02:05.490 And now I've built up my whole structure. 61 00:02:05.490 --> 00:02:08.170 I can just say JSON.GET enterprise again 62 00:02:08.170 --> 00:02:09.927 and get back the complete structure, 63 00:02:09.927 --> 00:02:13.620 "captain":"kirk","designation":"NCC-1701". 64 00:02:13.620 --> 00:02:14.670 Very cool stuff. 65 00:02:14.670 --> 00:02:16.340 It's just that simple. 66 00:02:16.340 --> 00:02:17.330 Now to actually use that, 67 00:02:17.330 --> 00:02:19.450 you could use the command line interface there, 68 00:02:19.450 --> 00:02:22.150 either directly or through a lower-level API 69 00:02:22.150 --> 00:02:24.120 like Redis-py. 70 00:02:24.120 --> 00:02:26.090 But there's also something I want to introduce now as well 71 00:02:26.090 --> 00:02:27.920 called Redis OM. 72 00:02:27.920 --> 00:02:30.140 And this is a high-level API to Redis 73 00:02:30.140 --> 00:02:33.290 that wraps the complexity of RedisJSON and RediSearch, 74 00:02:33.290 --> 00:02:34.840 so you don't really have to think about 75 00:02:34.840 --> 00:02:36.850 how it works under the hood at a low level. 76 00:02:36.850 --> 00:02:39.290 It stands for Redis Object Mapping. 77 00:02:39.290 --> 00:02:40.776 So it's a high-level API 78 00:02:40.776 --> 00:02:43.670 that just lets you deal with native data structures. 79 00:02:43.670 --> 00:02:45.450 So within Python, for example, 80 00:02:45.450 --> 00:02:47.720 I can just deal with Python classes 81 00:02:47.720 --> 00:02:49.870 and Redis OM will just sort of magically take care 82 00:02:49.870 --> 00:02:51.650 of storing it and querying that data 83 00:02:51.650 --> 00:02:54.350 with very simple commands like save or find 84 00:02:54.350 --> 00:02:55.310 or something like that. 85 00:02:55.310 --> 00:02:57.474 So it's a much easier way to interact 86 00:02:57.474 --> 00:03:00.933 with RedisJSON and RediSearch and even RedisCore. 87 00:03:02.440 --> 00:03:04.330 All the modules and the Redis commands under the hood 88 00:03:04.330 --> 00:03:05.200 are abstracted away. 89 00:03:05.200 --> 00:03:06.730 So you don't have to think about it. 90 00:03:06.730 --> 00:03:08.730 It's just like writing a normal script. 91 00:03:08.730 --> 00:03:11.440 You'll see how simple it is when we go into an example. 92 00:03:11.440 --> 00:03:13.077 It's available for several languages, 93 00:03:13.077 --> 00:03:16.730 .NET, Node.js, Python, and Spring. 94 00:03:16.730 --> 00:03:19.750 Since most of you that are my students are Python people, 95 00:03:19.750 --> 00:03:21.660 we're gonna stick with Python for our example here, 96 00:03:21.660 --> 00:03:25.350 but same concept applies to these other languages as well. 97 00:03:25.350 --> 00:03:26.710 Really easy to use, 98 00:03:26.710 --> 00:03:28.730 for example, here's how you would actually set up 99 00:03:28.730 --> 00:03:30.610 a nested data structure. 100 00:03:30.610 --> 00:03:31.750 This is coming from our activity 101 00:03:31.750 --> 00:03:33.500 that we're about to do, by the way. 102 00:03:33.500 --> 00:03:35.460 All you need to do is say from redis_om, 103 00:03:35.460 --> 00:03:36.690 if we're in Python, 104 00:03:36.690 --> 00:03:37.760 import the stuff we need, 105 00:03:37.760 --> 00:03:42.600 in this case, JsonModel and EmbeddedJsonModel and Field. 106 00:03:42.600 --> 00:03:44.100 So with EmbeddedJsonModel, 107 00:03:44.100 --> 00:03:46.200 we can embed one structure within another 108 00:03:46.200 --> 00:03:48.270 and Redis OM will just figure out 109 00:03:48.270 --> 00:03:51.410 how to make that work in JSON and with RedisJSON. 110 00:03:51.410 --> 00:03:54.308 We are then going to define a Product class here 111 00:03:54.308 --> 00:03:56.560 that contains a StockCode as a string, 112 00:03:56.560 --> 00:03:58.150 as Description as a string, and a UnitPrice, 113 00:03:58.150 --> 00:04:00.500 which is a floating point number in this example. 114 00:04:00.500 --> 00:04:03.160 And since it's deriving from EmbeddedJsonModel, 115 00:04:03.160 --> 00:04:06.460 it will automatically inherit all the Redis OM features 116 00:04:06.460 --> 00:04:08.790 for actually writing that out to Redis 117 00:04:08.790 --> 00:04:10.800 and reading from it or querying it. 118 00:04:10.800 --> 00:04:12.470 Because it's an EmbeddedJsonModel, 119 00:04:12.470 --> 00:04:15.040 I can embed it within another class here. 120 00:04:15.040 --> 00:04:17.260 So I'm setting up an Order class here 121 00:04:17.260 --> 00:04:19.330 that derives from just JsonModel. 122 00:04:19.330 --> 00:04:21.720 It contains an invoice number, which is a string, 123 00:04:21.720 --> 00:04:23.490 a Quantity, which is an integer, 124 00:04:23.490 --> 00:04:25.870 and here I'm saying I'm going to index that 125 00:04:25.870 --> 00:04:27.210 for searching later on. 126 00:04:27.210 --> 00:04:29.480 We'll talk about that more in a little bit. 127 00:04:29.480 --> 00:04:30.550 We're gonna have an InvoiceDate, 128 00:04:30.550 --> 00:04:32.719 that's just the datetime.date, a CustomerID. 129 00:04:32.719 --> 00:04:35.220 And here we're embedding that Product class 130 00:04:35.220 --> 00:04:36.650 and calling it Item. 131 00:04:36.650 --> 00:04:38.290 So that Item field there will expand 132 00:04:38.290 --> 00:04:40.380 into its own Product data structure here as well. 133 00:04:40.380 --> 00:04:41.213 And finally, 134 00:04:41.213 --> 00:04:43.740 we'll end up with a string-based Country code as well. 135 00:04:43.740 --> 00:04:45.790 So that's RedisJSON and Redis OM. 136 00:04:45.790 --> 00:04:48.640 We have enough now to get started and import some data 137 00:04:48.640 --> 00:04:50.610 just using Redis OM in a structured format. 138 00:04:50.610 --> 00:04:52.510 So let's dive in and make that happen.