1 00:00:00,180 --> 00:00:02,290 So a few additional lectures on DynamoDB 2 00:00:02,290 --> 00:00:04,180 because there are some more advanced questions now 3 00:00:04,180 --> 00:00:06,500 that are being asked of you at the exam. 4 00:00:06,500 --> 00:00:09,190 So the first feature I want to talk to you about is the TTL 5 00:00:09,190 --> 00:00:10,710 or Time to Live. 6 00:00:10,710 --> 00:00:13,290 The TTL means that you define a column 7 00:00:13,290 --> 00:00:16,160 and based on that column, the items that have 8 00:00:16,160 --> 00:00:19,270 an expiry date based on that column will expire 9 00:00:19,270 --> 00:00:20,810 and get deleted from the table. 10 00:00:20,810 --> 00:00:22,960 Hence the name, Time to Live. 11 00:00:22,960 --> 00:00:25,760 So the Time to Live feature in DynamoDB is provided to you 12 00:00:25,760 --> 00:00:28,340 at no extra cost and the deletions will not use 13 00:00:28,340 --> 00:00:31,270 any of the WCU or RCU. 14 00:00:31,270 --> 00:00:34,940 It's a background task and the DynamoDB service itself 15 00:00:34,940 --> 00:00:36,550 does it periodically 16 00:00:36,550 --> 00:00:39,040 and why would you use TTL, well maybe to reduce 17 00:00:39,040 --> 00:00:41,490 the storage and manage the table size over time 18 00:00:41,490 --> 00:00:43,850 because at some point, maybe you don't need a row 19 00:00:43,850 --> 00:00:46,580 to be in DynamoDB because it would have expired. 20 00:00:46,580 --> 00:00:49,370 It could also help you to adhere to regulatory norms. 21 00:00:49,370 --> 00:00:51,470 For example, if you need to keep (mumbles) 22 00:00:51,470 --> 00:00:53,620 only for seven days in some table. 23 00:00:53,620 --> 00:00:55,893 And TTL must be enabled per row so you define 24 00:00:55,893 --> 00:00:58,490 your TTL column as we'll see in the hands-on 25 00:00:58,490 --> 00:01:00,410 and we'll add a date there and we'll see 26 00:01:00,410 --> 00:01:01,670 in which format in a second. 27 00:01:01,670 --> 00:01:03,646 And then, on the pro basis, 28 00:01:03,646 --> 00:01:06,570 DynamoDB will delete these rows based on the expiration. 29 00:01:06,570 --> 00:01:09,030 So typically, it can happen pretty quickly but the guarantee 30 00:01:09,030 --> 00:01:12,490 is that DynamoDB will delete or will try to its best 31 00:01:12,490 --> 00:01:15,910 to delete a row within 48 hours of expiration, 32 00:01:15,910 --> 00:01:18,920 so it won't happen the second after the item expires 33 00:01:18,920 --> 00:01:21,420 but within 48 hours it should happen. 34 00:01:21,420 --> 00:01:24,410 The deleted items due to TTL will also be deleted 35 00:01:24,410 --> 00:01:28,020 from any indexes, so your GSI or LSI. 36 00:01:28,020 --> 00:01:31,160 And if you wanted to somehow recover these deleted items, 37 00:01:31,160 --> 00:01:34,000 we just saw what streams is, so DynamoDB streams 38 00:01:34,000 --> 00:01:36,000 upon a delete due to TTL. 39 00:01:36,000 --> 00:01:38,810 We'll have an event and maybe you can recover 40 00:01:38,810 --> 00:01:40,800 these expired items if you wanted to. 41 00:01:40,800 --> 00:01:41,880 So that's it for all the theory. 42 00:01:41,880 --> 00:01:44,580 Let's just go into the practice to see how TTL works. 43 00:01:44,580 --> 00:01:46,490 So let's go to creating a new table 44 00:01:46,490 --> 00:01:50,300 and I'll call it DemoTTL and I'll have user ID 45 00:01:50,300 --> 00:01:51,980 as my partition key. 46 00:01:51,980 --> 00:01:53,360 I will not use the default settings. 47 00:01:53,360 --> 00:01:54,470 I will choose provisioned. 48 00:01:54,470 --> 00:01:56,850 I will not have auto scaling and I will just have 49 00:01:56,850 --> 00:01:59,910 one read capacity unit and one write capacity unit. 50 00:01:59,910 --> 00:02:02,250 I'll choose the default for the encryption at rest 51 00:02:02,250 --> 00:02:03,420 and click on Create. 52 00:02:03,420 --> 00:02:04,770 Okay, so here is my table. 53 00:02:04,770 --> 00:02:05,923 It is being created. 54 00:02:06,820 --> 00:02:09,650 And now lets go to items and I'm going to create an item. 55 00:02:09,650 --> 00:02:12,720 So user ID, I'll say user_123. 56 00:02:12,720 --> 00:02:16,640 And maybe you'll have a name, it's Stephane. 57 00:02:16,640 --> 00:02:19,730 And here, I'm going to add an expire_on attribute. 58 00:02:19,730 --> 00:02:22,300 Now you can name this column whatever you want 59 00:02:22,300 --> 00:02:26,540 but it has to be a number and you'll set expire_on. 60 00:02:26,540 --> 00:02:28,255 And I will choose expire_on name 61 00:02:28,255 --> 00:02:31,320 because it's pretty obvious about what it does. 62 00:02:31,320 --> 00:02:34,920 So, now you do time to epoch on Google, 63 00:02:34,920 --> 00:02:36,960 and you have this website called epochconverter.com 64 00:02:36,960 --> 00:02:38,110 which is a great website. 65 00:02:38,110 --> 00:02:40,440 It allows you to take a time stamp, for example this one, 66 00:02:40,440 --> 00:02:43,620 and it gives you a number which represents an epoch time 67 00:02:43,620 --> 00:02:47,210 which is a number of milliseconds since the year '70. 68 00:02:47,210 --> 00:02:49,410 So, what we'll do is that we'll set okay, 69 00:02:49,410 --> 00:02:53,580 this one will expire at time 58, human to date timestamp 70 00:02:53,580 --> 00:02:55,730 and here's my epoch timestamp. 71 00:02:55,730 --> 00:02:58,640 And so what I'm going to do is copy this here. 72 00:02:58,640 --> 00:03:00,860 So this item should expire in five minutes. 73 00:03:00,860 --> 00:03:04,540 I'll create another item, maybe for user_234. 74 00:03:04,540 --> 00:03:09,540 And for its name, I will say Laura and the name is Laura. 75 00:03:10,430 --> 00:03:13,400 And then finally, again I will add an expire_on number 76 00:03:13,400 --> 00:03:16,380 so expire_on, and this time I'm going to select 77 00:03:16,380 --> 00:03:19,210 in one hour from now, so here we go. 78 00:03:19,210 --> 00:03:20,590 I paste this one here. 79 00:03:20,590 --> 00:03:21,660 Click on Save. 80 00:03:21,660 --> 00:03:25,180 And now I have two rows with two expire_on attributes. 81 00:03:25,180 --> 00:03:28,700 So I go to Overview, and I click on Time to Live attribute 82 00:03:28,700 --> 00:03:30,480 and I click on Manage TTL. 83 00:03:30,480 --> 00:03:33,010 Here, I will say that my column to look out 84 00:03:33,010 --> 00:03:35,440 for the TTL is expire_on 85 00:03:35,440 --> 00:03:38,300 and I could enable DynamoDB streams if I wanted 86 00:03:38,300 --> 00:03:41,300 to receive that as an update whenever it gets deleted. 87 00:03:41,300 --> 00:03:43,130 But here we can run a preview and say, 88 00:03:43,130 --> 00:03:46,100 show me the items that will expire by right now. 89 00:03:46,100 --> 00:03:49,250 If I run the preview, it says no items are matching. 90 00:03:49,250 --> 00:03:53,360 But if I go to a few minutes from now and do a run review 91 00:03:53,360 --> 00:03:55,110 it says I found one item 92 00:03:55,110 --> 00:03:57,960 this one that should expire by these dates. 93 00:03:57,960 --> 00:04:00,720 And if I go even further and run again the preview, 94 00:04:00,720 --> 00:04:03,020 now Laura is also going to be expired. 95 00:04:03,020 --> 00:04:05,260 So that TTL attribute just works fine 96 00:04:05,260 --> 00:04:06,480 and you can simulate here 97 00:04:06,480 --> 00:04:08,440 and when you're ready, click on Continue. 98 00:04:08,440 --> 00:04:11,230 And here we go, now it's being enabled. 99 00:04:11,230 --> 00:04:12,540 And so what will happen is that 100 00:04:12,540 --> 00:04:15,010 as we add items to this table, 101 00:04:15,010 --> 00:04:17,740 they will get deleted by DynamoDB in the background 102 00:04:17,740 --> 00:04:20,710 due to the TTL value on this column expire_on. 103 00:04:20,710 --> 00:04:21,990 So that's it, all you need to know. 104 00:04:21,990 --> 00:04:24,840 I hope that helps and I will see you in the next lecture.