WEBVTT 1 00:00:00.090 --> 00:00:00.960 In this lesson, 2 00:00:00.960 --> 00:00:02.730 we're going to talk about performance issues. 3 00:00:02.730 --> 00:00:05.460 Specifically, sluggish or slow performance 4 00:00:05.460 --> 00:00:08.010 that you're going to observe over time on a system. 5 00:00:08.010 --> 00:00:09.060 Now, performance issues 6 00:00:09.060 --> 00:00:10.470 are one of the hardest things 7 00:00:10.470 --> 00:00:13.470 to diagnose properly inside of a computer system 8 00:00:13.470 --> 00:00:15.330 because they could be hardware related, 9 00:00:15.330 --> 00:00:18.150 software related, or a mixture of the two. 10 00:00:18.150 --> 00:00:20.700 For this reason, it is really difficult to figure 11 00:00:20.700 --> 00:00:22.620 out exactly what the issue is, 12 00:00:22.620 --> 00:00:23.453 but I'm going to give you 13 00:00:23.453 --> 00:00:25.931 a couple of general guidelines in this lesson. 14 00:00:25.931 --> 00:00:28.650 Now, when you start to troubleshoot performance issues, 15 00:00:28.650 --> 00:00:30.720 you should always use a structured approach, 16 00:00:30.720 --> 00:00:32.160 and try to compartmentalize 17 00:00:32.160 --> 00:00:33.240 each of the different issues 18 00:00:33.240 --> 00:00:35.370 that could be causing that performance issue. 19 00:00:35.370 --> 00:00:36.300 So to begin, 20 00:00:36.300 --> 00:00:37.800 we need to know what a baseline is 21 00:00:37.800 --> 00:00:38.947 for our given system. 22 00:00:38.947 --> 00:00:42.000 For example, let's say that I have a system 23 00:00:42.000 --> 00:00:44.070 that has a three gigahertz processor, 24 00:00:44.070 --> 00:00:45.660 16 gigabytes of ram, 25 00:00:45.660 --> 00:00:47.460 a one terabyte hard drive, 26 00:00:47.460 --> 00:00:50.520 and a one gigabit per second networking card. 27 00:00:50.520 --> 00:00:53.070 Now, knowing all of those specifications is fine, 28 00:00:53.070 --> 00:00:54.420 but, if I've never actually 29 00:00:54.420 --> 00:00:56.550 touched that system before or looked at it, 30 00:00:56.550 --> 00:00:58.950 I don't really know how they're all working together. 31 00:00:58.950 --> 00:01:00.600 So when you have a new system, 32 00:01:00.600 --> 00:01:01.980 you want to observe that system 33 00:01:01.980 --> 00:01:03.720 and the way it currently operates. 34 00:01:03.720 --> 00:01:05.820 Taking the example system I just gave you, 35 00:01:05.820 --> 00:01:07.680 when we first get that brand new system, 36 00:01:07.680 --> 00:01:10.080 we should know what normal feels like. 37 00:01:10.080 --> 00:01:12.000 How fast does it really operate? 38 00:01:12.000 --> 00:01:14.340 How much RAM is it using on a daily basis 39 00:01:14.340 --> 00:01:15.957 when we have just the operating system loaded, 40 00:01:15.957 --> 00:01:19.290 or the operating system and an office suite open? 41 00:01:19.290 --> 00:01:21.390 We also need to know how fast is the network. 42 00:01:21.390 --> 00:01:23.280 Because just because we have a one gigabit 43 00:01:23.280 --> 00:01:24.270 per second card, 44 00:01:24.270 --> 00:01:25.770 doesn't mean we're getting one gigabit 45 00:01:25.770 --> 00:01:27.360 per second of throughput. 46 00:01:27.360 --> 00:01:28.470 In addition to all that, 47 00:01:28.470 --> 00:01:29.910 we have to look at the hard drive. 48 00:01:29.910 --> 00:01:31.860 I said it was a one terabyte hard drive, 49 00:01:31.860 --> 00:01:32.790 but I didn't tell you, 50 00:01:32.790 --> 00:01:34.290 was it a solid state drive? 51 00:01:34.290 --> 00:01:36.726 Was it a hard drive running at 5400 RPMs, 52 00:01:36.726 --> 00:01:39.480 or 7,200 RPMs, or 10,000 RPMs. 53 00:01:39.480 --> 00:01:42.000 And all of those are going to have different performance. 54 00:01:42.000 --> 00:01:42.870 Now, once you know 55 00:01:42.870 --> 00:01:44.670 what the normal performance is, 56 00:01:44.670 --> 00:01:45.630 you can then determine 57 00:01:45.630 --> 00:01:48.000 what is sluggish or slow performance, 58 00:01:48.000 --> 00:01:49.350 and then you'll be able to identify 59 00:01:49.350 --> 00:01:50.430 what those areas are, 60 00:01:50.430 --> 00:01:53.190 that you need to focus your troubleshooting efforts on. 61 00:01:53.190 --> 00:01:54.900 Now, in addition to being able to identify 62 00:01:54.900 --> 00:01:56.490 which of those subsystems 63 00:01:56.490 --> 00:01:58.080 is actually going to be the problem 64 00:01:58.080 --> 00:01:59.490 based on our sluggish performance 65 00:01:59.490 --> 00:02:00.450 that we're observing, 66 00:02:00.450 --> 00:02:02.130 we can then do configuration changes 67 00:02:02.130 --> 00:02:04.260 to optimize those subsystems. 68 00:02:04.260 --> 00:02:06.030 For example, if we start seeing 69 00:02:06.030 --> 00:02:07.110 that the average user 70 00:02:07.110 --> 00:02:09.150 is running out of RAM all the time, 71 00:02:09.150 --> 00:02:10.500 we might want to do an upgrade 72 00:02:10.500 --> 00:02:12.570 from eight gigabytes to 16 gigabytes, 73 00:02:12.570 --> 00:02:15.090 or 16 gigabytes to 32 gigabytes. 74 00:02:15.090 --> 00:02:16.320 Because again, this is one of the 75 00:02:16.320 --> 00:02:18.090 simplest and easiest things you can do 76 00:02:18.090 --> 00:02:19.648 to increase the performance of a system. 77 00:02:19.648 --> 00:02:22.380 But, if you are experiencing the sluggish performance, 78 00:02:22.380 --> 00:02:23.820 and you look at your free ram, 79 00:02:23.820 --> 00:02:26.550 and you have eight gigabytes of free ram out of 16, 80 00:02:26.550 --> 00:02:28.530 then upgrading to 32 is probably not 81 00:02:28.530 --> 00:02:30.630 going to do much for you in terms of performance, 82 00:02:30.630 --> 00:02:33.270 because you're not already using all the ram you have. 83 00:02:33.270 --> 00:02:34.320 And so these are the kind of things 84 00:02:34.320 --> 00:02:35.153 you have to think about 85 00:02:35.153 --> 00:02:37.290 when your troubleshooting performance issues. 86 00:02:37.290 --> 00:02:38.336 Now, in addition to all of that, 87 00:02:38.336 --> 00:02:39.630 you also want to make sure 88 00:02:39.630 --> 00:02:41.430 your system isn't overheating. 89 00:02:41.430 --> 00:02:42.471 A lot of our modern systems 90 00:02:42.471 --> 00:02:45.570 have throttling to their processing units 91 00:02:45.570 --> 00:02:47.790 to be able to protect themselves from overheating. 92 00:02:47.790 --> 00:02:50.010 So, if your system is starting to heat up, 93 00:02:50.010 --> 00:02:53.010 that processor that should be operating at three gigahertz, 94 00:02:53.010 --> 00:02:54.750 might actually downgrade itself 95 00:02:54.750 --> 00:02:57.180 to two gigahertz or 1.5 gigahertz 96 00:02:57.180 --> 00:02:59.430 to try to lower the thermal load that it's creating, 97 00:02:59.430 --> 00:03:01.740 and be able to help the system from overheating. 98 00:03:01.740 --> 00:03:03.270 Because, if a system overheats, 99 00:03:03.270 --> 00:03:05.190 it could reboot or it could shut down. 100 00:03:05.190 --> 00:03:07.200 So, a lot of your temperature sensors 101 00:03:07.200 --> 00:03:09.060 will actually slow down components 102 00:03:09.060 --> 00:03:10.860 such as the graphics processing unit, 103 00:03:10.860 --> 00:03:12.330 or the central processing unit 104 00:03:12.330 --> 00:03:14.492 to be able to lower the overall thermal load. 105 00:03:14.492 --> 00:03:16.363 If you have a faulty temperature sensor, 106 00:03:16.363 --> 00:03:17.670 this can actually indicate 107 00:03:17.670 --> 00:03:19.710 you have a higher temperature than really exists, 108 00:03:19.710 --> 00:03:21.630 and therefore, your CPU is going to start 109 00:03:21.630 --> 00:03:23.940 lowering its performance to compensate for that. 110 00:03:23.940 --> 00:03:25.429 So keep this in mind as well. 111 00:03:25.429 --> 00:03:26.790 Another thing to consider 112 00:03:26.790 --> 00:03:28.530 when you're troubleshooting performance issues, 113 00:03:28.530 --> 00:03:30.780 is any possible misconfigurations. 114 00:03:30.780 --> 00:03:33.420 For example, let's say you did a memory upgrade 115 00:03:33.420 --> 00:03:35.970 from eight gigabytes to 16 gigabytes. 116 00:03:35.970 --> 00:03:37.260 Well, when you did that, 117 00:03:37.260 --> 00:03:39.360 you may have taken out two four gigabyte modules, 118 00:03:39.360 --> 00:03:41.760 and put in two new eight gigabyte modules, 119 00:03:41.760 --> 00:03:44.250 but you put them into slots zero and one, 120 00:03:44.250 --> 00:03:46.320 instead of slots zero and two. 121 00:03:46.320 --> 00:03:49.440 And so your memory is not operating in dual channel mode. 122 00:03:49.440 --> 00:03:51.990 Instead, it's only operating in single channel mode. 123 00:03:51.990 --> 00:03:54.060 This will give you more sluggish performance 124 00:03:54.060 --> 00:03:54.990 than you may have had 125 00:03:54.990 --> 00:03:56.760 with the old eight gigabyte system, 126 00:03:56.760 --> 00:03:57.990 because you're only able to access 127 00:03:57.990 --> 00:03:59.130 64 bits at a time 128 00:03:59.130 --> 00:04:01.230 instead of 128 bits at a time, 129 00:04:01.230 --> 00:04:03.330 effectively cutting your performance in half 130 00:04:03.330 --> 00:04:05.190 even though you doubled your memory. 131 00:04:05.190 --> 00:04:07.200 So again, these type of misconfigurations 132 00:04:07.200 --> 00:04:10.530 can really have cascading effects throughout your system. 133 00:04:10.530 --> 00:04:11.970 Another common misconfiguration 134 00:04:11.970 --> 00:04:13.290 could be in the operating system 135 00:04:13.290 --> 00:04:15.166 or the application software itself. 136 00:04:15.166 --> 00:04:17.100 For example, I've seen a lot of people 137 00:04:17.100 --> 00:04:19.320 who increase the page size inside of windows, 138 00:04:19.320 --> 00:04:21.240 or the swap space inside of Linux. 139 00:04:21.240 --> 00:04:23.400 They think this is going to give them better performance, 140 00:04:23.400 --> 00:04:24.540 but instead, this is actually 141 00:04:24.540 --> 00:04:26.160 just giving you more virtual memory, 142 00:04:26.160 --> 00:04:27.630 which causes a lot more swapping 143 00:04:27.630 --> 00:04:29.400 from physical memory to your hard disc 144 00:04:29.400 --> 00:04:30.630 or solid state device, 145 00:04:30.630 --> 00:04:33.000 and thereby slows down your entire system. 146 00:04:33.000 --> 00:04:34.680 So you need to keep all these things in mind 147 00:04:34.680 --> 00:04:36.390 as you start configuring your system. 148 00:04:36.390 --> 00:04:39.240 And remember, none of these things work in isolation. 149 00:04:39.240 --> 00:04:40.707 When you have these performance issues, 150 00:04:40.707 --> 00:04:42.990 it can be a myriad of different things 151 00:04:42.990 --> 00:04:45.300 all working against you at the same time. 152 00:04:45.300 --> 00:04:46.620 It could be the operating system, 153 00:04:46.620 --> 00:04:47.850 it could be your applications, 154 00:04:47.850 --> 00:04:49.260 it could be your configurations, 155 00:04:49.260 --> 00:04:50.130 it could be your network, 156 00:04:50.130 --> 00:04:52.050 or it could be your hardware. 157 00:04:52.050 --> 00:04:53.400 And being able to identify this 158 00:04:53.400 --> 00:04:55.500 by breaking everything up into subsystems 159 00:04:55.500 --> 00:04:56.550 as you're considering it, 160 00:04:56.550 --> 00:04:58.290 and identifying those performance issues, 161 00:04:58.290 --> 00:04:59.550 is going to be critical for you 162 00:04:59.550 --> 00:05:01.233 as a technician in the field.