1 00:00:05,330 --> 00:00:07,470 Hello and welcome to this lecture on Cuba. 2 00:00:07,470 --> 00:00:08,600 It is controllers. 3 00:00:08,600 --> 00:00:10,300 My name is Mom child Mohamed. 4 00:00:10,540 --> 00:00:14,960 In this lecture we will discuss about Cuban it is controllers. 5 00:00:14,960 --> 00:00:17,900 Controllers are the brains behind Cuban interests. 6 00:00:17,900 --> 00:00:23,860 They are the processes that monitor Cuban natives objects and respond accordingly. 7 00:00:23,900 --> 00:00:33,090 In this lecture we will discuss about one controller in particular and that is the replication controller. 8 00:00:33,130 --> 00:00:38,350 So what is a replica and why do we need a replication controller. 9 00:00:38,350 --> 00:00:44,070 Let's go back to our first scenario where we had a single pod running our application. 10 00:00:44,170 --> 00:00:48,270 What if for some reason our application crashes and the pod fails. 11 00:00:48,460 --> 00:00:56,050 Users will no longer be able to access our application to prevent users from losing access to our application. 12 00:00:56,050 --> 00:01:01,380 We would like to have more than one instance or pod running at the same time. 13 00:01:01,450 --> 00:01:06,660 That way if one failed we still have our application running on the other one. 14 00:01:06,820 --> 00:01:13,870 The replication controller helps us run multiple instances of a single pod in the cabinet as cluster 15 00:01:14,350 --> 00:01:22,090 thus providing high availability so does that mean you can't use a replication controller if you plan 16 00:01:22,090 --> 00:01:24,260 to have a single pod. 17 00:01:24,310 --> 00:01:25,120 No. 18 00:01:25,180 --> 00:01:32,590 Even if you have a single pod the replication controller can help by automatically bringing up a new 19 00:01:32,590 --> 00:01:35,240 pod when the existing one fails. 20 00:01:35,500 --> 00:01:42,370 Thus the replication controller ensures that the specified number of parts are running at all times 21 00:01:42,520 --> 00:01:51,060 even if it's just one or 100 another reason we need a replication controller is to create multiple paths 22 00:01:51,120 --> 00:01:53,400 to share the load across them. 23 00:01:53,400 --> 00:02:01,080 For example in this simple scenario we have a single pod serving a set of users when the number of users 24 00:02:01,110 --> 00:02:01,860 increase. 25 00:02:01,860 --> 00:02:09,560 We deploy additional pod to balance the load across the two parts if the demand further increases and 26 00:02:09,560 --> 00:02:16,340 if we were to run out of resources on the first node we could deploy additional pods across the other 27 00:02:16,340 --> 00:02:18,240 nodes in the cluster. 28 00:02:18,320 --> 00:02:24,880 As you can see the replication controller spans across multiple nodes in the cluster. 29 00:02:24,950 --> 00:02:31,910 It helps us balance the load across multiple paths on different nodes as well as scale our application. 30 00:02:32,030 --> 00:02:41,040 When the demand increases it's important to note that there are two similar terms replication controller 31 00:02:41,190 --> 00:02:43,000 and replay cassette. 32 00:02:43,140 --> 00:02:50,550 Both have the same purpose but they're not the same replication controller is the older technology that 33 00:02:50,550 --> 00:02:58,280 is being replaced by replicate set properly has said is the new recommended way to setup replication. 34 00:02:58,350 --> 00:03:05,350 However whatever we discussed in the previous few slides remain applicable to both these technologies. 35 00:03:05,640 --> 00:03:11,050 There are minor differences in the way each works and we will look at that in a bit. 36 00:03:11,190 --> 00:03:18,170 As such we will try to stick to replicable sets in all of our demos and implementations going forward. 37 00:03:19,360 --> 00:03:23,760 Let us now look at how we create a replication controller. 38 00:03:23,770 --> 00:03:30,490 As with the previous lecture we start by creating a replication controller definition file we will name 39 00:03:30,490 --> 00:03:37,120 it our C dash to finish thought Jamil as with any Cuban Interest definition file. 40 00:03:37,130 --> 00:03:47,090 We have four sections the API version kind metadata and spec the API version is specific to what we 41 00:03:47,090 --> 00:03:53,990 are creating in this case replication controller is supported in Coburn it is API version V1. 42 00:03:54,710 --> 00:04:03,740 So we will set it as v1 the kind as we know is a replication controller under a metadata we will add 43 00:04:03,800 --> 00:04:12,500 a name and we will call it my ab dash RC and we will also add a few labels app and type and assign values 44 00:04:12,500 --> 00:04:12,860 to them. 45 00:04:13,820 --> 00:04:19,420 So far it has been very similar to how we created a pod in the previous section. 46 00:04:19,460 --> 00:04:26,510 The next is the most crucial part of the definition file and that is the specification written as spec 47 00:04:27,420 --> 00:04:29,980 for any Cuban it is definition file. 48 00:04:30,030 --> 00:04:34,710 The spec section defines what's inside the object we are creating. 49 00:04:34,710 --> 00:04:43,130 In this case we know that the replication controller creates multiple instances of a part but what part 50 00:04:43,860 --> 00:04:51,330 we create a template section under spec to provide a part template to be used by the replication controller 51 00:04:51,480 --> 00:04:57,440 to create replicas now how do we define the part template. 52 00:04:57,440 --> 00:05:03,250 It's not that hard because we have already done that in the previous exercise. 53 00:05:03,290 --> 00:05:08,430 Remember we created a pod definition file in the previous exercise. 54 00:05:08,540 --> 00:05:16,140 We could reuse the contents of the file to populate the template section move all the contents of the 55 00:05:16,140 --> 00:05:22,680 pod definition file into the template section of the replication controller except for the first few 56 00:05:22,680 --> 00:05:31,060 lines which are API version and kind remember whatever we move must be under the template section meaning 57 00:05:31,060 --> 00:05:35,980 this should be intended to the right and have more spaces before them. 58 00:05:36,040 --> 00:05:44,070 Then the template line itself they should be children of the template section looking at our file. 59 00:05:44,080 --> 00:05:47,890 Now we now have two method metadata sections. 60 00:05:47,890 --> 00:05:57,550 One is for the replication controller and another for the pod and we have two aspects sections one for 61 00:05:57,550 --> 00:06:02,180 each. 62 00:06:02,350 --> 00:06:05,540 We have nested to definition files together. 63 00:06:05,740 --> 00:06:11,750 The replication controller being the parent and the part definition being the child. 64 00:06:11,880 --> 00:06:14,300 Now there is something still missing. 65 00:06:14,300 --> 00:06:20,340 We haven't mentioned how many replicas we need in the replication controller for that. 66 00:06:20,340 --> 00:06:27,330 Add another property to the spec called replicas and input the number of replicas you need under it. 67 00:06:27,380 --> 00:06:34,970 Remember that the template and replicas are direct children of spec sections so they are siblings and 68 00:06:34,970 --> 00:06:42,590 must be on the same vertical line which means having equal number of spaces before them once the file 69 00:06:42,590 --> 00:06:50,720 is ready run the cubes GTL create command and input the file using the dash f parameter the replication 70 00:06:50,720 --> 00:06:57,110 controller is created when the replication controller is created it first creates the part using the 71 00:06:57,110 --> 00:07:00,920 part definition template as many as required which is 3. 72 00:07:00,920 --> 00:07:08,810 In this case to view the list of created replication controllers run the cube CDL get replication controller 73 00:07:08,810 --> 00:07:15,170 command and you will see the replication controller listed we can also see the desired number of friendly 74 00:07:15,180 --> 00:07:20,750 cars or parts the current number of friendly cars and how many of them are ready in the output. 75 00:07:21,110 --> 00:07:26,620 If you would like to see the parts that were created by the replication controller run the cube CDL 76 00:07:26,780 --> 00:07:30,760 get parts command and you will see three parts running. 77 00:07:30,770 --> 00:07:37,580 Note that all of them are starting with the name of the replication controller which is my ab dash RC 78 00:07:37,970 --> 00:07:45,920 indicating that they are all created automatically by the replication controller what we just saw was 79 00:07:45,980 --> 00:07:52,670 the replication controller let us now look at replica set it is very similar to replication controller 80 00:07:53,120 --> 00:08:00,950 as usual first we have API version kind metadata and spec the API version though is a bit different 81 00:08:01,190 --> 00:08:07,550 it is apps slash to be one which is different from what we had before for application controller which 82 00:08:07,550 --> 00:08:15,200 was just V1 if you get this wrong you are likely to get an error that looks like this it would say no 83 00:08:15,200 --> 00:08:22,220 match for kind replica set because the specified carbonates API version has no support for replica set 84 00:08:23,890 --> 00:08:29,730 the kind would be replicas set and we add name and labels in the metadata 85 00:08:33,220 --> 00:08:39,140 the specification section looks very similar to replication controller it has a template section where 86 00:08:39,140 --> 00:08:42,230 we provide part definition as before. 87 00:08:42,500 --> 00:08:47,750 So I'm going to copy the contents over from part definition file and we have number of replicas which 88 00:08:47,750 --> 00:08:49,340 is set to three. 89 00:08:49,400 --> 00:08:56,660 However there is one major difference between replication controller and replica set has set requires 90 00:08:56,750 --> 00:09:06,610 a selector definition the selector section helps the replica set identify what parts fall under it but 91 00:09:06,610 --> 00:09:10,510 why would you have to specify what parts fall under it. 92 00:09:10,600 --> 00:09:18,170 If we have provided the contents of the part definition file itself in the template it's because replicas 93 00:09:18,180 --> 00:09:24,790 set can also manage parts that were not created as part of the replica at creation. 94 00:09:24,920 --> 00:09:32,360 Say for example the reports created before the creation of the replica set that match labels specified 95 00:09:32,360 --> 00:09:33,480 in the selector. 96 00:09:33,680 --> 00:09:39,010 The replica set will also take those parts into consideration when creating the replicas. 97 00:09:39,930 --> 00:09:46,170 I will elaborate this in the next slide but before we get into that I would like to mention that the 98 00:09:46,170 --> 00:09:52,770 selector is one of the major differences between replication controller and replica set the selector 99 00:09:52,800 --> 00:09:59,100 is not a required field in case of a replication controller but it is still available when you skip 100 00:09:59,100 --> 00:10:01,560 it as we did in the previous slide. 101 00:10:01,620 --> 00:10:08,910 It assumes it to be the same as the labels provided in the part definition file in case of replica said 102 00:10:09,120 --> 00:10:16,890 a user input is required for this property and it has to be written in the form of match labels as shown 103 00:10:16,890 --> 00:10:17,780 here. 104 00:10:17,970 --> 00:10:24,960 The match labeled selector simply matches the labels specified under it to the labels on the part. 105 00:10:24,960 --> 00:10:32,100 The replica set selector also provides many other options for matching labels that were not available 106 00:10:32,130 --> 00:10:33,060 in the replication. 107 00:10:33,060 --> 00:10:37,970 Controller and as always to create a replica set. 108 00:10:38,000 --> 00:10:45,020 Run the cube CTF and create command providing the definition file as input and to see the created replicas 109 00:10:45,080 --> 00:10:53,540 run the cubes TTL get replica set command to get list of parts simply run the cube control get parts. 110 00:10:53,540 --> 00:10:55,920 Command. 111 00:10:56,130 --> 00:10:59,620 So what is the deal with labels and selectors. 112 00:10:59,620 --> 00:11:03,060 Why do we label our parts and objects in corporate entities. 113 00:11:03,190 --> 00:11:05,510 Let us look at a simple scenario. 114 00:11:05,770 --> 00:11:12,660 Say we deployed three instances of our front end web application as three part. 115 00:11:12,760 --> 00:11:18,610 We would like to create a replication controller or replica set to ensure that we have three active 116 00:11:18,610 --> 00:11:20,770 parts at any time. 117 00:11:20,770 --> 00:11:24,300 And yes that is one of the use case of replica sets. 118 00:11:24,400 --> 00:11:27,460 You can use it to monitor existing parts. 119 00:11:27,490 --> 00:11:34,610 If you have them already created as cities in this example in case they were not created the replica 120 00:11:34,610 --> 00:11:37,190 set will create them for you. 121 00:11:37,310 --> 00:11:44,000 The role of the replica set is to monitor the parts and if any of them were to fail deploy new ones. 122 00:11:44,000 --> 00:11:49,570 The replica set is in fact a process that monitors the parts. 123 00:11:49,610 --> 00:11:51,050 Now how does the replica set. 124 00:11:51,230 --> 00:11:51,980 No. 125 00:11:52,010 --> 00:11:53,960 What part to monitor. 126 00:11:53,960 --> 00:11:59,450 There could be hundreds of other parts in the cluster running different applications. 127 00:11:59,450 --> 00:12:08,280 This is where labelling our parts during creation comes in handy we could now provide these labels as 128 00:12:08,280 --> 00:12:15,810 a filter for replica set under the selector section we use to match labels filter and provide the same 129 00:12:15,810 --> 00:12:18,980 label that we used while creating the parts. 130 00:12:19,140 --> 00:12:27,670 This way the replica set knows which parts to monitor the same concept of labels and selectors is used 131 00:12:27,760 --> 00:12:30,880 in many other places throughout Cuba. 132 00:12:30,880 --> 00:12:39,380 Notice now let me ask you a question along the same lines in the replica set specification section. 133 00:12:39,380 --> 00:12:45,150 We learned that there are three sections template replicas and the selector. 134 00:12:45,290 --> 00:12:50,990 We need three replicas and we have updated our selector based on our discussion in the previous slide 135 00:12:52,050 --> 00:12:58,260 say for instance we have the same scenario as in the previous slide where we have three existing parts 136 00:12:58,350 --> 00:13:05,100 that were created already and we need to create a replica set to monitor the parts to ensure there are 137 00:13:05,190 --> 00:13:10,860 a minimum of three running at all times when the replication controller is created. 138 00:13:10,890 --> 00:13:19,300 It is not going to deploy a new instance of part as three of them with matching labels are already created. 139 00:13:19,480 --> 00:13:26,320 In that case do we really need to provide a template section in the type replica set specification. 140 00:13:26,320 --> 00:13:32,960 Since we are not expecting the replica set to create a new port on deployment yes we do. 141 00:13:33,070 --> 00:13:40,180 Because in case one of the parts were to fail in the future the replicas that needs to create a new 142 00:13:40,180 --> 00:13:47,230 one to maintain the desired number of pods and for the replica set to create a new pod the template 143 00:13:47,260 --> 00:13:55,020 definition section is required let's now look at how we scaled the replica set. 144 00:13:55,260 --> 00:14:00,470 Say we started with three replicas in the future we decided to scale to six. 145 00:14:00,600 --> 00:14:05,310 How do we update our replica set to scale to six replicas. 146 00:14:05,310 --> 00:14:07,420 Well there are multiple ways to do it. 147 00:14:07,440 --> 00:14:12,660 The first is to update the number of replicas in the definition file to 6 148 00:14:15,340 --> 00:14:22,920 then run the cubes TTL replace command to specify the same file using the dash f parameter and that 149 00:14:22,920 --> 00:14:26,130 will update the replica set to have six replicas. 150 00:14:27,270 --> 00:14:31,890 The second way to do it is to run the cube control scale command. 151 00:14:32,080 --> 00:14:38,280 Use the replicas parameter to provide the new number of replicas and specify the same file as input 152 00:14:39,570 --> 00:14:47,870 human either input the definition file or provide the Republicans at name in the type name format however 153 00:14:47,990 --> 00:14:55,610 remember that using the file name as input will not result in the number of replicas being updated automatically 154 00:14:55,610 --> 00:14:56,960 in the file. 155 00:14:56,960 --> 00:15:03,410 In other words the number of friendly cars in the replica set definition file will still be three even 156 00:15:03,410 --> 00:15:09,260 though you scaled your replica set to have six replicas using the cube control scale command and the 157 00:15:09,260 --> 00:15:11,580 file as input. 158 00:15:11,590 --> 00:15:18,370 There are also options available for automatically scaling the replica set based on load but that is 159 00:15:18,370 --> 00:15:22,080 an advanced topic and we will discuss it at a later time. 160 00:15:23,280 --> 00:15:29,480 Let us now review the commands real quick the cube control create command as we know is used to create 161 00:15:29,480 --> 00:15:36,440 a replica set or basically any object in Cuba and it is depending on the file we are providing as input. 162 00:15:36,920 --> 00:15:44,130 You must provide the input file using the dash f parameter use the cube control get command to see List 163 00:15:44,130 --> 00:15:50,010 of replica sets created used to keep control delete replica set command followed by the name of the 164 00:15:50,010 --> 00:15:56,340 replica set to delete the replica set and then we have the cube controlled replace Command to replace 165 00:15:56,340 --> 00:16:03,480 or update the replica set and also the cube control scale command scale replica set simply from the 166 00:16:03,480 --> 00:16:09,020 command line without having to modify the file that's it for this lecture.