1 00:00:00,570 --> 00:00:01,580 Okay, so now let's talk 2 00:00:01,580 --> 00:00:03,660 about EC2 Instance Types. 3 00:00:03,660 --> 00:00:06,190 So there are different types of EC2 instances 4 00:00:06,190 --> 00:00:08,189 that you can use for different use cases 5 00:00:08,189 --> 00:00:10,490 and they have different types of optimization. 6 00:00:10,490 --> 00:00:11,980 And let's go check out this link 7 00:00:11,980 --> 00:00:13,990 and we'll see we have for now, 8 00:00:13,990 --> 00:00:17,150 seven different types of EC2 instances. 9 00:00:17,150 --> 00:00:19,290 So this website on the AWS website 10 00:00:19,290 --> 00:00:21,000 is what we're interested into. 11 00:00:21,000 --> 00:00:24,110 And as we can see, we have different types of instances. 12 00:00:24,110 --> 00:00:25,680 We have general purpose, compute, optimized, 13 00:00:25,680 --> 00:00:27,150 memory optimize and so on. 14 00:00:27,150 --> 00:00:29,020 And so for each type of instance 15 00:00:29,020 --> 00:00:30,860 we have different families. 16 00:00:30,860 --> 00:00:33,110 And so as we can see this website is going to 17 00:00:33,110 --> 00:00:37,540 be the reference for us to look at Institute Instance types 18 00:00:37,540 --> 00:00:39,630 and know their costs and other specificity. 19 00:00:39,630 --> 00:00:41,130 What I'm going to do though, is just walk you 20 00:00:41,130 --> 00:00:44,537 through a high-level overview of how they work in AWS. 21 00:00:45,678 --> 00:00:48,130 AWS will have the following naming convention. 22 00:00:48,130 --> 00:00:49,230 For example, we'll be talking 23 00:00:49,230 --> 00:00:52,520 about an M five dot two X large type of instance. 24 00:00:52,520 --> 00:00:53,640 What does that mean? 25 00:00:53,640 --> 00:00:56,870 Well, M is going to be called the Instance Class. 26 00:00:56,870 --> 00:00:59,120 Okay. And this is going to be, for example, in this case 27 00:00:59,120 --> 00:01:01,190 a general purpose type of instance, 28 00:01:01,190 --> 00:01:04,060 five is generation of the instance. 29 00:01:04,060 --> 00:01:06,454 So as AWS improves the hardware over time 30 00:01:06,454 --> 00:01:09,480 if we release a new generation of hardware, and so 31 00:01:09,480 --> 00:01:13,130 after M five, if they improve the M type of instance class 32 00:01:13,130 --> 00:01:15,770 then we'll go to M six 33 00:01:15,770 --> 00:01:19,040 and then finally the two X large represented the size 34 00:01:19,040 --> 00:01:20,390 within the instance class. 35 00:01:20,390 --> 00:01:23,480 So, it starts as small and then large 36 00:01:23,480 --> 00:01:25,250 and then two X large four X large and so on. 37 00:01:25,250 --> 00:01:27,210 So it represents the size of the instance, 38 00:01:27,210 --> 00:01:29,840 and the more the size, the more the memory 39 00:01:29,840 --> 00:01:32,560 the more the CPU is going to have on your instance. 40 00:01:32,560 --> 00:01:34,760 So from an exam perspective, what do you need to know? 41 00:01:34,760 --> 00:01:36,620 Well, we'll talk about a few different types 42 00:01:36,620 --> 00:01:37,453 of instance types. 43 00:01:37,453 --> 00:01:38,580 So you have a general purpose 44 00:01:38,580 --> 00:01:41,020 and these are great for diversity of workloads 45 00:01:41,020 --> 00:01:44,340 such as web servers or code repositories. 46 00:01:44,340 --> 00:01:45,460 They will have a good balance 47 00:01:45,460 --> 00:01:48,300 between compute, memory, networking. 48 00:01:48,300 --> 00:01:49,210 And so in this course 49 00:01:49,210 --> 00:01:51,170 we'll be using general purpose instances. 50 00:01:51,170 --> 00:01:52,730 We'll be using the T2 micro 51 00:01:52,730 --> 00:01:56,930 which is the free tier, general purpose type of instance. 52 00:01:56,930 --> 00:01:58,450 On the website that I just showed you 53 00:01:58,450 --> 00:02:00,340 you will see all the different types 54 00:02:00,340 --> 00:02:02,460 of instance that our general purpose 55 00:02:02,460 --> 00:02:04,010 and this is going to evolve over time, 56 00:02:04,010 --> 00:02:05,310 So I won't update this slide. 57 00:02:05,310 --> 00:02:06,380 But you can always refer back 58 00:02:06,380 --> 00:02:09,680 to the AWS website to check what the instances are 59 00:02:09,680 --> 00:02:12,540 in the general purpose type of family. 60 00:02:12,540 --> 00:02:14,720 Then we have compute optimized, 61 00:02:14,720 --> 00:02:16,920 and these are instances are great, 62 00:02:16,920 --> 00:02:19,830 and optimized for compute intensive tasks. 63 00:02:19,830 --> 00:02:23,110 So what requires a high level of processor? 64 00:02:23,110 --> 00:02:24,290 Well, if, for example, it could be 65 00:02:24,290 --> 00:02:26,210 if you're batch processing some data 66 00:02:26,210 --> 00:02:27,980 if you're doing media transcoding 67 00:02:27,980 --> 00:02:30,020 if you need a high-performance web servers 68 00:02:30,020 --> 00:02:32,810 if you're doing high performance, computing is called HPC. 69 00:02:32,810 --> 00:02:34,240 If you're doing machine learning 70 00:02:34,240 --> 00:02:36,720 or if you have a dedicated gaming server. 71 00:02:36,720 --> 00:02:40,900 So all these things are tasks that require a very good CPU 72 00:02:40,900 --> 00:02:42,190 very good compute side. 73 00:02:42,190 --> 00:02:43,990 And so Ec2, 74 00:02:43,990 --> 00:02:47,860 instances do have this kind of particularity 75 00:02:47,860 --> 00:02:51,110 and for now all the computer optimized instances 76 00:02:51,110 --> 00:02:55,090 in EC2, are of the C names. 77 00:02:55,090 --> 00:02:57,620 So C5, C6, and so on. 78 00:02:57,620 --> 00:03:01,530 Next, we have some EC2 instances that are memory optimized 79 00:03:01,530 --> 00:03:02,860 and there are going to be 80 00:03:02,860 --> 00:03:05,360 have a really fast performance for the type 81 00:03:05,360 --> 00:03:09,160 of workloads that will process large datasets in memory. 82 00:03:09,160 --> 00:03:10,410 So memory means RAM. 83 00:03:10,410 --> 00:03:12,360 And so the use cases are this is going to 84 00:03:12,360 --> 00:03:14,570 be high performance for relational or 85 00:03:14,570 --> 00:03:16,930 non-relational databases are mostly going to be 86 00:03:16,930 --> 00:03:18,790 in memory databases, 87 00:03:18,790 --> 00:03:20,730 distributed web-scale cache store. 88 00:03:20,730 --> 00:03:22,890 So for our elastic cache, for example 89 00:03:22,890 --> 00:03:24,770 in memory databases that are optimized 90 00:03:24,770 --> 00:03:27,370 for business intelligence or BI. 91 00:03:27,370 --> 00:03:30,130 And applications performing real-time processing 92 00:03:30,130 --> 00:03:32,850 of big unstructured data. 93 00:03:32,850 --> 00:03:36,170 So in terms of the names for the memory optimized instances 94 00:03:36,170 --> 00:03:39,690 there's going to be the R series because R stands for RAM 95 00:03:39,690 --> 00:03:42,680 but there's also going to be X one high memory and Z one, 96 00:03:42,680 --> 00:03:44,230 but again, you don't have to remember the name 97 00:03:44,230 --> 00:03:47,450 of the instances, but good to know at a high level. 98 00:03:47,450 --> 00:03:51,700 Okay. And finally we'll have storage optimized instances. 99 00:03:51,700 --> 00:03:54,300 They're great when you are accessing a lot 100 00:03:54,300 --> 00:03:57,370 of data sets on the local storage. 101 00:03:57,370 --> 00:03:59,840 And so the use cases for storage optimized instances 102 00:03:59,840 --> 00:04:02,130 are going to be high-frequency online 103 00:04:02,130 --> 00:04:04,420 transactional processing, so OLTP systems. 104 00:04:04,420 --> 00:04:06,590 Relational and NoSQL databases. 105 00:04:06,590 --> 00:04:07,770 And we'll see those in details. 106 00:04:07,770 --> 00:04:09,490 When we get to the database sections. 107 00:04:09,490 --> 00:04:11,340 Cache for in-memory databases, for example, 108 00:04:11,340 --> 00:04:14,270 Reddit's data warehousing application 109 00:04:14,270 --> 00:04:15,930 distributed file systems 110 00:04:15,930 --> 00:04:19,070 and the search optimized instances in AWS 111 00:04:19,070 --> 00:04:21,790 will start with an I, a G, or H one. 112 00:04:21,790 --> 00:04:23,700 Okay. But again, don't have to remember this. 113 00:04:23,700 --> 00:04:25,530 I'm just giving you examples. 114 00:04:25,530 --> 00:04:27,060 So what does it mean? 115 00:04:27,060 --> 00:04:28,940 Let's compare a few instance types. 116 00:04:28,940 --> 00:04:30,470 So for example, for t2.micro 117 00:04:30,470 --> 00:04:34,980 we have one VCPU and one memory, one gigabytes of memory. 118 00:04:34,980 --> 00:04:38,830 And if you look for example, at r5.16xlarge 119 00:04:38,830 --> 00:04:43,330 we have 16 VCPU and 512 gigabytes of memory. 120 00:04:43,330 --> 00:04:46,330 So we can see there's a lot of more emphasis on the memory. 121 00:04:46,330 --> 00:04:49,137 If we compare it to example, to a c5d.4Xlarge 122 00:04:50,800 --> 00:04:54,700 we can see we have 16 VCPU and 32 gigabytes of memory. 123 00:04:54,700 --> 00:04:57,010 So less memory, more CPU 124 00:04:57,010 --> 00:04:59,180 and so on different network performance 125 00:04:59,180 --> 00:05:00,870 different EBS bandwidth and so on. 126 00:05:00,870 --> 00:05:03,260 So just to give you a point of comparison, 127 00:05:03,260 --> 00:05:05,810 and because we're using t2.micro in this course 128 00:05:05,810 --> 00:05:07,680 it is part of the AWS feature. 129 00:05:07,680 --> 00:05:11,520 So we get up to 750 hours per month of t2.micro. 130 00:05:11,520 --> 00:05:14,190 And if you want a website to compare all the 131 00:05:14,190 --> 00:05:15,720 EC2 Institute instances together, 132 00:05:15,720 --> 00:05:16,660 there's one that I really like, 133 00:05:16,660 --> 00:05:18,890 it's called instituteinstance.info, 134 00:05:18,890 --> 00:05:20,970 and I'll show it to you right now. 135 00:05:20,970 --> 00:05:23,480 So I am on the instituteinstances.info website 136 00:05:23,480 --> 00:05:24,520 and as we can see, 137 00:05:24,520 --> 00:05:28,480 we have a list of all the instances available in AWS. 138 00:05:28,480 --> 00:05:29,313 So really, a lot. 139 00:05:29,313 --> 00:05:31,060 We also get some information around 140 00:05:31,060 --> 00:05:33,200 the Linux on demand cost and an X reserves cost, 141 00:05:33,200 --> 00:05:35,140 and so on, so some cost information. 142 00:05:35,140 --> 00:05:38,430 We get information around the memory the number of VCPU. 143 00:05:38,430 --> 00:05:40,480 We can order by name, we can search it. 144 00:05:40,480 --> 00:05:41,610 So it's, it's quite handy. 145 00:05:41,610 --> 00:05:42,760 And I really like this website. 146 00:05:42,760 --> 00:05:44,930 And if you go and use AWS 147 00:05:44,930 --> 00:05:47,290 you probably will use this website as well. 148 00:05:47,290 --> 00:05:48,710 So that's it for this lecture. 149 00:05:48,710 --> 00:05:49,543 I hope you liked it. 150 00:05:49,543 --> 00:05:51,400 And I will see you in the next lecture.