WEBVTT 00:00:00.000 --> 00:00:10.920 Alright folks, so now we're going to begin our Encore architecture preparation. 00:00:10.920 --> 00:00:16.240 So we spent a little bit of time in day one or session number one looking a little bit 00:00:16.240 --> 00:00:23.360 at the architecture, the access layer, distribution layer, the core layer, but there are a few 00:00:23.360 --> 00:00:29.920 other aspects that we need to grasp to get ready for the certification exam when it comes 00:00:29.920 --> 00:00:37.160 to our architecture, such as stateful switchover, non-stop forwarding, our control plane, our 00:00:37.160 --> 00:00:38.160 data plane. 00:00:38.160 --> 00:00:45.040 We'll even have a discussion here on virtual servers and virtual switches. 00:00:45.040 --> 00:00:51.680 So let's begin here by taking a look at stateful switchover and non-stop forwarding. 00:00:51.680 --> 00:00:59.040 So first and foremost, stateful switchover, SSO, so if you see that on the exam, SSO, 00:00:59.040 --> 00:01:07.240 we're talking about stateful switchover, and if you see NSF, non-stop forwarding, that's 00:01:07.240 --> 00:01:10.320 what they're referring to in that particular case. 00:01:10.320 --> 00:01:20.680 So remember from day one that we indicated that it's important to have redundancy, and 00:01:20.680 --> 00:01:26.920 we are going to have redundancy in our access layer, our distribution layer, our core layer, 00:01:26.920 --> 00:01:31.760 our campus backbone, and that's going to come through redundant devices, it's going 00:01:31.760 --> 00:01:38.480 to come from redundant links, but let's talk about redundancy within those devices specifically 00:01:38.480 --> 00:01:40.480 at this point in time. 00:01:40.480 --> 00:01:50.320 What we have to remember is that our devices have a processing aspect associated with them. 00:01:50.320 --> 00:01:55.920 So for example, they have the ability to send spanning tree protocol messages. 00:01:55.920 --> 00:02:00.100 They have the ability to send dynamic routing protocol messages. 00:02:00.100 --> 00:02:04.720 They have the ability to think and make decisions. 00:02:04.720 --> 00:02:10.880 So with that in mind, I want you to consider the fact that the brains of the device, this 00:02:10.880 --> 00:02:18.340 processing capability of the device could fail, and if that fails, then our device goes 00:02:18.340 --> 00:02:20.460 down. 00:02:20.460 --> 00:02:26.780 So we need redundancy in the device, and in this particular case, we're talking about 00:02:26.780 --> 00:02:29.100 the supervisor module. 00:02:29.100 --> 00:02:36.100 Now the supervisor module would be the actual module that goes in the actual device, and 00:02:36.100 --> 00:02:39.820 that module would have something known as our route processor. 00:02:39.820 --> 00:02:43.060 So the route processor would be like your CPU as an example. 00:02:43.060 --> 00:02:48.700 So on the exam, you might see RP for route processor, or you'll see route processor. 00:02:48.700 --> 00:02:52.860 They might call it the RP module or the route processor module, or they might call it the 00:02:52.860 --> 00:02:56.940 supervisor module or just the supervisor. 00:02:56.940 --> 00:03:01.780 So keep in mind that if you see supervisor, supervisor module, route processor, route 00:03:01.780 --> 00:03:08.020 processor module, RP module, they're really referring to the same thing in that particular 00:03:08.020 --> 00:03:09.420 case. 00:03:09.420 --> 00:03:18.040 So we're dealing with something that provides control plane functions, tasks like making 00:03:18.040 --> 00:03:24.600 routing decisions, making management decisions, processing our protocols like spanning sheet 00:03:24.600 --> 00:03:36.480 protocols, CDP, LLDP, LACP, PEGP, DTP, VTP, EIGRP, OSPF, BGP, and so on and so forth. 00:03:36.480 --> 00:03:41.520 And these supervisor modules are typically found in our higher end equipment, so our 00:03:41.520 --> 00:03:49.080 modular type of equipment, and it's doing the control functions for us. 00:03:49.080 --> 00:03:55.880 So when it comes to the supervisor module, think CPU, the brains of the actual device, 00:03:55.880 --> 00:04:03.280 storing and running the operating system, so the Cisco iOS, the iOS XE, controlling 00:04:03.280 --> 00:04:09.360 what is going to be done for loading and managing and upgrading the actual software on the device. 00:04:09.360 --> 00:04:14.720 It's going to be able to manage and control the hardware components of our device as well, 00:04:14.720 --> 00:04:18.220 our line cards, our power supplies, our fans. 00:04:18.220 --> 00:04:24.540 It can monitor the status of a multitude of components and take actions depending on things 00:04:24.540 --> 00:04:26.320 like hardware failures. 00:04:26.320 --> 00:04:31.360 It even has the interface we need to configure and manage our device, whether it's the command 00:04:31.360 --> 00:04:34.920 line interface we're talking about or the graphical user interface we're talking about. 00:04:34.920 --> 00:04:38.520 It's all part of this supervisor module. 00:04:38.520 --> 00:04:43.160 So as you can see, the supervisor module is really, really important. 00:04:43.160 --> 00:04:49.160 And if our supervisor module ever fails, then we are out of luck. 00:04:49.160 --> 00:04:55.560 Our device is not going to be able to exchange the protocols it needs to, to ultimately build 00:04:55.560 --> 00:04:57.020 the tables we need. 00:04:57.020 --> 00:05:01.000 So it can ultimately make the forwarding decisions that it needs to make. 00:05:01.000 --> 00:05:09.120 So it's really important that when we are dealing with these modular devices, we spend 00:05:09.120 --> 00:05:15.600 some extra money and ensure that we have redundancy in regards to the brains, the control plane, 00:05:15.600 --> 00:05:19.320 the supervisor module of our devices. 00:05:19.320 --> 00:05:25.040 And what you're going to have here is you're going to have now high availability with the 00:05:25.040 --> 00:05:26.840 actual control plane. 00:05:26.840 --> 00:05:33.320 So if one module fails, then the backup can take over without disrupting operations. 00:05:33.320 --> 00:05:41.240 But in order for this to work properly, we need what is called stateful switchover, SSO. 00:05:41.240 --> 00:05:48.540 So stateful switchover, what we have is an active and passive scenario. 00:05:48.540 --> 00:05:51.760 So we have our two supervisor modules. 00:05:51.760 --> 00:05:56.480 One's going to be active and one's going to be passive inside of our chassis. 00:05:57.120 --> 00:06:01.760 And this is going to give us the chance to have that uninterrupted functionality that 00:06:01.760 --> 00:06:07.240 we need if one of those supervisor modules fails. 00:06:07.240 --> 00:06:10.920 So the active one is going to take care of all of the operations. 00:06:10.920 --> 00:06:16.600 The second one is going to be synchronized with the first one, folks, right? 00:06:16.600 --> 00:06:18.480 So active, passive. 00:06:18.480 --> 00:06:22.840 One is doing all the work and the other one is simply synchronized. 00:06:22.840 --> 00:06:26.840 So the standby is synchronized with the active. 00:06:26.840 --> 00:06:33.680 So they have the exact same configuration, but one's active, one's standby. 00:06:33.680 --> 00:06:41.640 So if the active supervisor module fails, the standby one can take care of what needs 00:06:41.640 --> 00:06:46.480 to be taken care of at that point in time without interruption, ensuring that there 00:06:46.480 --> 00:06:54.720 is minimal or hopefully no downtime when our supervisor module or our route processor, 00:06:54.720 --> 00:06:59.240 our RP, fails. 00:06:59.240 --> 00:07:03.080 Now what we need to realize is that we're talking about the control plane at this particular 00:07:03.080 --> 00:07:05.640 point in time. 00:07:05.640 --> 00:07:09.600 I now want to focus on nonstop forwarding. 00:07:09.600 --> 00:07:10.600 What is this about? 00:07:10.880 --> 00:07:16.560 This is actually about our data and our data plane, whereas SSO is more about the control 00:07:16.560 --> 00:07:17.560 plane. 00:07:17.560 --> 00:07:18.560 Not more about. 00:07:18.560 --> 00:07:21.160 It is the control plane. 00:07:21.160 --> 00:07:24.320 Think about our network traffic. 00:07:24.320 --> 00:07:34.760 Our network traffic, we want to continue flowing through our device. 00:07:34.760 --> 00:07:37.080 So network traffic's coming in. 00:07:37.080 --> 00:07:39.280 We want the network traffic to keep going. 00:07:39.280 --> 00:07:45.120 Remember, the network traffic is not going to be flowing through our route processor, 00:07:45.120 --> 00:07:46.120 our supervisor module. 00:07:46.120 --> 00:07:47.120 Right? 00:07:47.120 --> 00:07:48.120 This is the brains. 00:07:48.120 --> 00:07:54.560 We're going to have line cards in our device, and we want those line cards to continue doing 00:07:54.560 --> 00:07:55.560 what? 00:07:55.560 --> 00:07:59.320 Continue forwarding. 00:07:59.320 --> 00:08:07.560 So our network traffic needs to keep flowing, and this is where NSF comes into play, nonstop 00:08:07.560 --> 00:08:08.560 forwarding. 00:08:08.640 --> 00:08:11.200 It works together with SSO. 00:08:11.200 --> 00:08:12.200 Remember that. 00:08:12.200 --> 00:08:15.200 It works together with SSO. 00:08:15.200 --> 00:08:16.720 So you have SSO enabled. 00:08:16.720 --> 00:08:19.560 You have nonstop forwarding enabled. 00:08:19.560 --> 00:08:24.920 Nonstop forwarding needs SSO, but this nonstop forwarding provides what? 00:08:24.920 --> 00:08:34.020 It provides Layer 3 uninterrupted network slash routing operations when we have a failure 00:08:34.020 --> 00:08:35.800 of the supervisor modules. 00:08:35.800 --> 00:08:39.520 So think of protocols like BGP, OSPF, EIGRP. 00:08:39.520 --> 00:08:44.640 Those are NSF nonstop forwarding capable protocols. 00:08:44.640 --> 00:08:47.480 They're able to detect failure, react accordingly. 00:08:47.480 --> 00:08:54.160 So there's no operational interruption, and in a nutshell, what we're saying here is that 00:08:54.160 --> 00:09:03.600 when we have SSO with nonstop forwarding enabled, then the supervisor module can fail, and while 00:09:03.600 --> 00:09:10.720 the switchover is happening, traffic can continue to flow. 00:09:10.720 --> 00:09:14.960 So think about a protocol like OSPF. 00:09:14.960 --> 00:09:24.120 OSPF has a hello timer of 30, 30, no it doesn't, 10 seconds, and a dead timer of what? 00:09:24.120 --> 00:09:27.320 Four times the hello, so 40 seconds. 00:09:27.320 --> 00:09:37.680 So as long as the stateful switchover finishes within what, 40 seconds, then our line cards 00:09:37.680 --> 00:09:47.080 can continue to forward based on the information that they previously got, because the neighborship 00:09:47.080 --> 00:09:51.120 is still up on the other device, right? 00:09:51.120 --> 00:09:52.200 Think about it for a second. 00:09:52.200 --> 00:09:58.680 If our route processor fails, then the device on the other side would stop receiving what? 00:09:58.680 --> 00:09:59.680 Hello packets. 00:09:59.680 --> 00:10:04.280 Therefore, if it stops receiving hello packets within 40 seconds, it's going to tear down 00:10:04.280 --> 00:10:06.240 the neighborship. 00:10:06.240 --> 00:10:12.680 So with nonstop forwarding enabled, it ensures the device that has the failure keeps forwarding 00:10:12.680 --> 00:10:19.160 packets, and as long as the stateful switchover finishes within the dead timer setting, so 00:10:19.160 --> 00:10:24.680 40 seconds, well then the other neighbor has no idea we had a problem over here, because 00:10:24.680 --> 00:10:29.080 everything just keeps going well, and then stateful switchover finishes, and then those 00:10:29.080 --> 00:10:32.480 hello packets are getting exchanged, and everything's good. 00:10:32.480 --> 00:10:38.000 The neighboring device has no idea we were affected, or they could have been affected. 00:10:38.000 --> 00:10:39.000 Why? 00:10:39.000 --> 00:10:45.240 The high availability through redundancy, using things like what? 00:10:45.800 --> 00:10:51.840 Supervisor modules, with stateful switchover, and nonstop forwarding enabled. 00:10:51.840 --> 00:10:56.760 So line cards can continue forwarding data traffic with nonstop forwarding. 00:10:56.760 --> 00:11:03.160 All right, so keep in mind for the exam, if they are focusing the question around redundancy 00:11:03.160 --> 00:11:07.600 within a chassis, then they're really talking about the supervisor modules in that particular 00:11:07.600 --> 00:11:08.600 question. 00:11:08.600 --> 00:11:13.000 So they could be listed as route processor, RP, supervisor modules, but now pay attention 00:11:13.000 --> 00:11:14.000 to the details. 00:11:14.000 --> 00:11:20.320 If they're talking about control plane, yeah, and you're looking for answer like SSO, stateful 00:11:20.320 --> 00:11:26.720 switchover, but if they're talking about network traffic, data plane, traffic flowing through 00:11:26.720 --> 00:11:33.880 the device, then they are looking for nonstop forwarding as the answer option, all right? 00:11:33.880 --> 00:11:35.600 That's what I want you to take away from this discussion. 00:11:35.600 --> 00:11:37.760 So make sure you know what a supervisor module is. 00:11:37.760 --> 00:11:41.320 Make sure you know you can have redundant supervisor modules in our modular chassis. 00:11:41.320 --> 00:11:46.200 Make sure you realize what stateful switchover is and what nonstop forwarding is to get yourself 00:11:46.200 --> 00:11:47.920 ready for the certification exam.