1 00:00:00,800 --> 00:00:02,330 Whether we're talking about children 2 00:00:02,330 --> 00:00:05,470 or full grown adults, no one enjoys 3 00:00:05,470 --> 00:00:07,930 having their stuff messed with. 4 00:00:07,930 --> 00:00:09,910 In this lesson, we are going to talk about 5 00:00:09,910 --> 00:00:13,550 how to handle interruptions in Azure Stream Analytics. 6 00:00:13,550 --> 00:00:15,980 And this may be one of the shortest lessons 7 00:00:15,980 --> 00:00:17,740 that I have done to date. 8 00:00:17,740 --> 00:00:21,053 In this lesson, we're going to be discussing paired regions. 9 00:00:22,080 --> 00:00:25,640 Yes, no, that's it. Just paired regions, so pretty fast. 10 00:00:25,640 --> 00:00:27,360 With that, let's go ahead and just jump in 11 00:00:27,360 --> 00:00:28,710 and start talking about it. 12 00:00:30,020 --> 00:00:31,730 Before we talk about paired regions, 13 00:00:31,730 --> 00:00:34,970 we first have to start talking about availability sets. 14 00:00:34,970 --> 00:00:39,640 So an availability set is just a separation of your data 15 00:00:40,660 --> 00:00:43,180 in separate hardware within a datacenter. 16 00:00:43,180 --> 00:00:44,930 So think 2 different computers 17 00:00:44,930 --> 00:00:48,490 both storing your data within the same datacenter. 18 00:00:48,490 --> 00:00:51,110 So this is going to give you some protection 19 00:00:51,110 --> 00:00:53,740 if one machine fails, okay? 20 00:00:53,740 --> 00:00:55,160 It's not going to protect you 21 00:00:55,160 --> 00:00:57,100 if the datacenter fails, though. 22 00:00:57,100 --> 00:01:01,230 If that's the case, you got to look at an availability zone. 23 00:01:01,230 --> 00:01:04,210 So think about Chicago having 2 different datacenters. 24 00:01:04,210 --> 00:01:08,220 Both of those datacenters are storing your data 25 00:01:08,220 --> 00:01:11,840 and they both have multiple copies of your data 26 00:01:11,840 --> 00:01:14,020 within those availability sets. 27 00:01:14,020 --> 00:01:15,890 However, we still have a problem 28 00:01:15,890 --> 00:01:19,600 because what if a tornado comes and just wipes out Chicago? 29 00:01:19,600 --> 00:01:21,280 Well, if that's your concern, 30 00:01:21,280 --> 00:01:23,730 then we need to look at regions. 31 00:01:23,730 --> 00:01:25,370 And when we talk about regions, 32 00:01:25,370 --> 00:01:28,800 now what we're talking about is the separation of location. 33 00:01:28,800 --> 00:01:31,890 So I can have a datacenter in Chicago 34 00:01:31,890 --> 00:01:34,360 and then I can have a datacenter in New York 35 00:01:34,360 --> 00:01:37,530 and that would be a region pair 36 00:01:37,530 --> 00:01:40,360 of East U.S. and Central U.S. 37 00:01:40,360 --> 00:01:42,880 So I have 2 different datacenters 38 00:01:42,880 --> 00:01:45,170 in 2 different regions. 39 00:01:45,170 --> 00:01:48,420 So now I have separation of the region, 40 00:01:48,420 --> 00:01:50,050 I have separation of the datacenter, 41 00:01:50,050 --> 00:01:52,640 and I have separation of the data within the datacenter 42 00:01:52,640 --> 00:01:55,160 with those availability sets. 43 00:01:55,160 --> 00:01:57,220 Now, I mentioned that pair. 44 00:01:57,220 --> 00:01:58,940 So regions pair together 45 00:01:58,940 --> 00:02:02,430 and Azure has 2 different regions that pair together. 46 00:02:02,430 --> 00:02:06,300 So for instance, like I said, you can have a Central U.S. 47 00:02:06,300 --> 00:02:08,240 and an East U.S. or you can have 48 00:02:08,240 --> 00:02:10,050 an East U.S. and a West U.S. 49 00:02:10,050 --> 00:02:12,403 And so you choose a pair. 50 00:02:14,300 --> 00:02:15,690 And then within that, 51 00:02:15,690 --> 00:02:19,470 that is going to be within the much larger geography. 52 00:02:19,470 --> 00:02:22,573 So U.S., that would be your geography. 53 00:02:23,840 --> 00:02:27,370 And so when we look at building protection, 54 00:02:27,370 --> 00:02:31,500 Azure recommends that if you have mission-critical jobs 55 00:02:31,500 --> 00:02:36,140 that you need to create an availability on a paired region. 56 00:02:36,140 --> 00:02:38,670 So, basically, what you're doing is you are going to look 57 00:02:38,670 --> 00:02:41,980 at a region pair, like Central and East U.S., 58 00:02:41,980 --> 00:02:43,570 and then what you're going to do is 59 00:02:43,570 --> 00:02:45,940 you are going to run identical jobs 60 00:02:45,940 --> 00:02:48,330 in both of those paired regions. 61 00:02:48,330 --> 00:02:49,830 That way, if something happens, 62 00:02:49,830 --> 00:02:54,210 you have some protection in case of a critical failure. 63 00:02:54,210 --> 00:02:57,490 Now, of course, Microsoft's going to work very hard 64 00:02:57,490 --> 00:02:59,270 to avoid having a failure. 65 00:02:59,270 --> 00:03:01,740 But if you have something that is mission critical, 66 00:03:01,740 --> 00:03:03,140 Microsoft is recommending 67 00:03:03,140 --> 00:03:05,983 that you use an Azure paired region. 68 00:03:08,320 --> 00:03:10,663 So with that, let's talk about the review. 69 00:03:11,530 --> 00:03:14,840 So first, the Stream Analytics guarantee. 70 00:03:14,840 --> 00:03:17,920 So Microsoft promises that when they do updates, 71 00:03:17,920 --> 00:03:21,240 they will update in separate batches for paired regions. 72 00:03:21,240 --> 00:03:23,630 That means that East U.S. and West U.S. 73 00:03:23,630 --> 00:03:25,930 are not going to be updated at the same time. 74 00:03:25,930 --> 00:03:28,480 They're going to update one and then update the other. 75 00:03:29,770 --> 00:03:32,540 So with that, the deployment of the updates 76 00:03:32,540 --> 00:03:34,030 don't occur at the same time. 77 00:03:34,030 --> 00:03:37,060 So because of that, make sure that if you need 78 00:03:37,060 --> 00:03:40,400 to run a mission-critical stream analytics job 79 00:03:40,400 --> 00:03:42,610 that you're going to deploy that 80 00:03:42,610 --> 00:03:45,610 to 2 different paired regions at the same time. 81 00:03:45,610 --> 00:03:47,620 If you do that, you're going to be well-protected 82 00:03:47,620 --> 00:03:50,030 against any mayhem that might occur. 83 00:03:50,030 --> 00:03:52,350 So with that, that's it, we're done. 84 00:03:52,350 --> 00:03:54,270 We're ready to move on to the next lesson. 85 00:03:54,270 --> 00:03:55,570 Just make sure you keep that 86 00:03:55,570 --> 00:03:57,603 Stream Analytics guarantee in mind.