WEBVTT 1 00:00:00.120 --> 00:00:01.800 In this lesson, we're going to talk about 2 00:00:01.800 --> 00:00:03.330 drive performance issues, 3 00:00:03.330 --> 00:00:06.300 specifically, about the smart diagnostic tool 4 00:00:06.300 --> 00:00:07.740 and a metric known as 5 00:00:07.740 --> 00:00:10.320 the input output operations per second. 6 00:00:10.320 --> 00:00:12.630 First, let's talk about SMART. 7 00:00:12.630 --> 00:00:14.970 Now, SMART is an acronym and SMART stands for 8 00:00:14.970 --> 00:00:18.090 the Self-Monitoring Analysis and Reporting Technology. 9 00:00:18.090 --> 00:00:19.683 And it's built into most hard disk drives 10 00:00:19.683 --> 00:00:22.320 that you're going to be using on a daily basis. 11 00:00:22.320 --> 00:00:24.180 Now, this smart technology is actually 12 00:00:24.180 --> 00:00:26.640 a self-diagnostic program that's going to alert 13 00:00:26.640 --> 00:00:28.410 the operating system if it detects 14 00:00:28.410 --> 00:00:29.790 that there's a failure that's occurred 15 00:00:29.790 --> 00:00:32.070 or if a failure is eminent. 16 00:00:32.070 --> 00:00:33.870 Now, the smart system has been around 17 00:00:33.870 --> 00:00:35.760 for about 20 to 30 years, 18 00:00:35.760 --> 00:00:37.620 and it started out in hard disk drives, 19 00:00:37.620 --> 00:00:40.890 but has also been migrated into solid state devices. 20 00:00:40.890 --> 00:00:43.740 This SMART technology is something that's going to compliment 21 00:00:43.740 --> 00:00:46.260 your operating system, your BIOS, your UEFI 22 00:00:46.260 --> 00:00:48.150 and other systems like that. 23 00:00:48.150 --> 00:00:49.650 The whole idea with SMART is 24 00:00:49.650 --> 00:00:51.660 that it can then monitor your hard drive 25 00:00:51.660 --> 00:00:53.430 and be able to understand the health status 26 00:00:53.430 --> 00:00:55.230 of that particular drive. 27 00:00:55.230 --> 00:00:57.420 As it looks at it, it's going to be looking at things like 28 00:00:57.420 --> 00:01:00.180 the temperature of the drive, the drives overall health, 29 00:01:00.180 --> 00:01:01.590 how fast it's spinning if you're dealing 30 00:01:01.590 --> 00:01:02.700 with a hard disk drive, 31 00:01:02.700 --> 00:01:04.290 how many bad blocks there are if you're dealing 32 00:01:04.290 --> 00:01:07.230 with a solid-state device and things like that. 33 00:01:07.230 --> 00:01:09.270 Now, the important thing to remember with SMART is 34 00:01:09.270 --> 00:01:12.360 that it's not designed to keep your drives healthy. 35 00:01:12.360 --> 00:01:14.460 Instead, it's designed to identify 36 00:01:14.460 --> 00:01:15.840 when they're having problems 37 00:01:15.840 --> 00:01:18.090 and when they might be failing in the future. 38 00:01:18.090 --> 00:01:19.410 So keep this in mind. 39 00:01:19.410 --> 00:01:22.650 Their goal here is not to be able to fix your drive 40 00:01:22.650 --> 00:01:24.690 but simply to monitor and figure out 41 00:01:24.690 --> 00:01:26.643 when something bad is about to happen. 42 00:01:27.480 --> 00:01:29.190 This self-monitoring, analysis 43 00:01:29.190 --> 00:01:31.620 and reporting technology is designed to measure lots 44 00:01:31.620 --> 00:01:33.810 of different attributes about your hard drive 45 00:01:33.810 --> 00:01:35.610 and your solid-state devices. 46 00:01:35.610 --> 00:01:37.470 For example, it can monitor things like 47 00:01:37.470 --> 00:01:40.290 your read error rate, the spin-up time of the hard disc, 48 00:01:40.290 --> 00:01:43.530 the reallocated sectors count, the seek error rate, 49 00:01:43.530 --> 00:01:46.230 the power-on hours, high temperature alerts, 50 00:01:46.230 --> 00:01:47.850 and other things like this. 51 00:01:47.850 --> 00:01:50.370 By monitoring all these different pieces of information, 52 00:01:50.370 --> 00:01:53.340 the SMART utility is able to notify the operating system 53 00:01:53.340 --> 00:01:54.600 of any pending issues 54 00:01:54.600 --> 00:01:56.910 or problems that may exist in the future. 55 00:01:56.910 --> 00:01:58.650 Now, one of the metrics that the SMART utility 56 00:01:58.650 --> 00:02:01.920 is monitoring is known as your input output operations 57 00:02:01.920 --> 00:02:03.900 per second or IOPS. 58 00:02:03.900 --> 00:02:06.000 Now, IOPS is a really important metric 59 00:02:06.000 --> 00:02:07.950 to be able to look at on a system. 60 00:02:07.950 --> 00:02:10.800 Because as you're measuring how fast things can be inputted 61 00:02:10.800 --> 00:02:12.660 or outputted to a storage device, 62 00:02:12.660 --> 00:02:14.940 this is one of the key metrics for performance 63 00:02:14.940 --> 00:02:17.820 that you need to understand when dealing with storage. 64 00:02:17.820 --> 00:02:19.680 Now, if you're comparing a hard disk drive 65 00:02:19.680 --> 00:02:21.270 to a solid-state device, 66 00:02:21.270 --> 00:02:23.820 generally you're going to have much lower IOPS 67 00:02:23.820 --> 00:02:26.970 on a hard disk drive than you will on a solid-state device. 68 00:02:26.970 --> 00:02:28.800 This is because a solid-state device 69 00:02:28.800 --> 00:02:31.410 has much faster seek times because it doesn't have to move 70 00:02:31.410 --> 00:02:34.380 a platter around to be able to read that information, 71 00:02:34.380 --> 00:02:36.600 but that is not always the case. 72 00:02:36.600 --> 00:02:38.130 There are some very inexpensive 73 00:02:38.130 --> 00:02:42.450 and cheaply made SSDs that have much lower IOPS counts 74 00:02:42.450 --> 00:02:45.270 than do some really fast hard disk drives. 75 00:02:45.270 --> 00:02:48.060 So you have to remember that not everything is clear cut 76 00:02:48.060 --> 00:02:50.070 where solid-state devices are always faster 77 00:02:50.070 --> 00:02:52.860 than hard drives, but that is a good general rule. 78 00:02:52.860 --> 00:02:54.450 So keep this in mind. 79 00:02:54.450 --> 00:02:55.950 If you're working in the cloud, 80 00:02:55.950 --> 00:02:58.710 you don't actually have access to the underlying storage. 81 00:02:58.710 --> 00:03:01.260 So the only way you can measure performance is 82 00:03:01.260 --> 00:03:02.670 by measuring the IOPS, 83 00:03:02.670 --> 00:03:04.830 which is the input output operations per second, 84 00:03:04.830 --> 00:03:07.350 because you don't know if they're using hard disk drives 85 00:03:07.350 --> 00:03:10.260 or solid-state devices underneath that storage layer, 86 00:03:10.260 --> 00:03:12.210 because it's being abstracted from you 87 00:03:12.210 --> 00:03:14.880 when you're using virtualization and cloud computing. 88 00:03:14.880 --> 00:03:17.610 And this is why we like to look at the IOPS number 89 00:03:17.610 --> 00:03:20.070 or that input output operations per second, 90 00:03:20.070 --> 00:03:22.740 because this is the best measure of performance we have 91 00:03:22.740 --> 00:03:24.630 when we are reading and writing data 92 00:03:24.630 --> 00:03:26.700 to a particular storage device. 93 00:03:26.700 --> 00:03:28.860 Now, if you start seeing that your overall performance 94 00:03:28.860 --> 00:03:31.500 is going down when you're talking about your storage devices 95 00:03:31.500 --> 00:03:32.850 and you measure your IOPS 96 00:03:32.850 --> 00:03:35.160 and you see that number is going down as well, 97 00:03:35.160 --> 00:03:37.200 this indicates you're having some sort of 98 00:03:37.200 --> 00:03:40.620 a bottleneck reading and writing to those storage devices. 99 00:03:40.620 --> 00:03:42.120 This could be an issue with the hardware 100 00:03:42.120 --> 00:03:44.250 or it could be an issue with the software. 101 00:03:44.250 --> 00:03:46.530 You'd have to look into that and be able to determine that 102 00:03:46.530 --> 00:03:48.523 by using tools like the SMARt utilities 103 00:03:48.523 --> 00:03:51.030 to be able to determine is the drive reading 104 00:03:51.030 --> 00:03:52.440 and writing at the right speed, 105 00:03:52.440 --> 00:03:55.320 or is it the operating system that's actually slowing down 106 00:03:55.320 --> 00:03:56.790 and not able to read and write 107 00:03:56.790 --> 00:03:59.040 to the device as fast as it needs to? 108 00:03:59.040 --> 00:04:00.720 And this is why it's important to look at things 109 00:04:00.720 --> 00:04:03.990 holistically because a low input output operations 110 00:04:03.990 --> 00:04:06.570 per second could be an issue with the hardware, 111 00:04:06.570 --> 00:04:08.850 or it could be something with the application 112 00:04:08.850 --> 00:04:11.580 or the general system resources that are being overloaded 113 00:04:11.580 --> 00:04:13.440 because of a lot of other things that are happening 114 00:04:13.440 --> 00:04:15.270 on that particular system. 115 00:04:15.270 --> 00:04:16.920 If you're experiencing slow performance 116 00:04:16.920 --> 00:04:18.210 when it comes to storage, 117 00:04:18.210 --> 00:04:19.350 some of the things you want to look at 118 00:04:19.350 --> 00:04:21.000 on a hard disk drive would be 119 00:04:21.000 --> 00:04:23.310 whether or not the files have become fragmented. 120 00:04:23.310 --> 00:04:24.150 And if they have, 121 00:04:24.150 --> 00:04:26.880 you're going to want to run a defragmentation tool. 122 00:04:26.880 --> 00:04:29.070 This will be able to take all the pieces of those files 123 00:04:29.070 --> 00:04:30.450 and put them back together 124 00:04:30.450 --> 00:04:34.020 which will then reduce your read right times to that device. 125 00:04:34.020 --> 00:04:36.720 This is because when you're writing to a hard disk drive, 126 00:04:36.720 --> 00:04:38.250 it's going to start taking that file 127 00:04:38.250 --> 00:04:39.083 that you're trying to write, 128 00:04:39.083 --> 00:04:41.430 and it will break it up into smaller pieces. 129 00:04:41.430 --> 00:04:43.860 Each of those pieces will be put onto the hard disk 130 00:04:43.860 --> 00:04:45.180 starting at the beginning 131 00:04:45.180 --> 00:04:47.910 and moving towards the end in any available space. 132 00:04:47.910 --> 00:04:49.800 So I want you to think about a hard disk drive 133 00:04:49.800 --> 00:04:51.300 like you would a book. 134 00:04:51.300 --> 00:04:54.630 Let's say I gave you a blank book that has 200 pages in it. 135 00:04:54.630 --> 00:04:57.780 That's all the storage you have, but you have a pencil, 136 00:04:57.780 --> 00:04:59.190 and as you're writing in the book, 137 00:04:59.190 --> 00:05:01.140 you start on page one and you keep writing 138 00:05:01.140 --> 00:05:02.670 until you get to page 50. 139 00:05:02.670 --> 00:05:04.320 Now, you went back to page three 140 00:05:04.320 --> 00:05:06.630 and decided you didn't need stuff on that page. 141 00:05:06.630 --> 00:05:07.800 So you erase that. 142 00:05:07.800 --> 00:05:09.780 Now you have some empty space in that book. 143 00:05:09.780 --> 00:05:12.840 You did the same thing on page seven and 12 and 15, 144 00:05:12.840 --> 00:05:13.740 and you went through the book 145 00:05:13.740 --> 00:05:15.330 and erase things you didn't need. 146 00:05:15.330 --> 00:05:17.190 Now when you want to write something else down, 147 00:05:17.190 --> 00:05:19.200 you're not going to start on page 51, 148 00:05:19.200 --> 00:05:21.660 but you're going to go back to page three where you erased that 149 00:05:21.660 --> 00:05:23.070 and fill in the blank. 150 00:05:23.070 --> 00:05:24.450 Now, once you run out of space, 151 00:05:24.450 --> 00:05:26.250 you're going to move to the next page that has blanks, 152 00:05:26.250 --> 00:05:27.420 and you're going to keep doing that 153 00:05:27.420 --> 00:05:30.000 until you get all the way back up to page 51, 154 00:05:30.000 --> 00:05:31.710 and then you'll keep writing again. 155 00:05:31.710 --> 00:05:33.480 That's how a hard drive works. 156 00:05:33.480 --> 00:05:35.640 So if you have a lot of little files 157 00:05:35.640 --> 00:05:36.630 that you've been deleting, 158 00:05:36.630 --> 00:05:39.270 you now have a lot of little holes all over your drive. 159 00:05:39.270 --> 00:05:41.130 And the next time you go to write something, 160 00:05:41.130 --> 00:05:43.050 it's going to start filling in those holes. 161 00:05:43.050 --> 00:05:45.540 But because you have to seek to those different places 162 00:05:45.540 --> 00:05:47.400 to gain that whole file again, 163 00:05:47.400 --> 00:05:49.950 this can lead to excessive read write times. 164 00:05:49.950 --> 00:05:52.410 So to be able to bring that time back down, 165 00:05:52.410 --> 00:05:54.510 you want to do what's called defragmentation. 166 00:05:54.510 --> 00:05:57.300 Where it goes through and puts all those files back together 167 00:05:57.300 --> 00:05:59.550 in the right order on the same pages. 168 00:05:59.550 --> 00:06:01.350 And it can move things around the drive 169 00:06:01.350 --> 00:06:03.180 and shuffle them around until everything 170 00:06:03.180 --> 00:06:04.860 is nice and orderly again. 171 00:06:04.860 --> 00:06:07.710 Over time though, that drive is going to get jumbled up again 172 00:06:07.710 --> 00:06:09.120 because of all of this deleting 173 00:06:09.120 --> 00:06:10.530 and rewriting that's occurring, 174 00:06:10.530 --> 00:06:12.810 and that's where fragmentation happens. 175 00:06:12.810 --> 00:06:14.730 As your drive becomes more fragmented, 176 00:06:14.730 --> 00:06:16.620 this means you're going to have lower performance 177 00:06:16.620 --> 00:06:19.890 when you're trying to seek or read data from that drive. 178 00:06:19.890 --> 00:06:22.770 Now, luckily, as you move to a solid-state device, 179 00:06:22.770 --> 00:06:24.720 fragmentation is no longer an issue, 180 00:06:24.720 --> 00:06:26.640 because we can access everything on that drive 181 00:06:26.640 --> 00:06:27.930 at any given time, 182 00:06:27.930 --> 00:06:29.250 because we don't have to move the platter 183 00:06:29.250 --> 00:06:30.870 to be able to access a particular part 184 00:06:30.870 --> 00:06:32.940 of that drive and get the data back off. 185 00:06:32.940 --> 00:06:35.490 So fragmentation is no longer an issue 186 00:06:35.490 --> 00:06:36.570 with solid-state devices, 187 00:06:36.570 --> 00:06:39.090 and you do not need to run defragmentation utilities 188 00:06:39.090 --> 00:06:40.530 on a solid-state device. 189 00:06:40.530 --> 00:06:43.230 But you do need to do it on a hard disk drive. 190 00:06:43.230 --> 00:06:45.480 Now, on the other hand, with a solid-state device, 191 00:06:45.480 --> 00:06:46.980 if you start running out of space 192 00:06:46.980 --> 00:06:48.780 and you have decreased capacity, 193 00:06:48.780 --> 00:06:50.070 you are going to notice that the drive 194 00:06:50.070 --> 00:06:51.990 is going to start slowing down on you. 195 00:06:51.990 --> 00:06:53.670 To be able to speed the drive back up, 196 00:06:53.670 --> 00:06:54.960 you're going to have to move some of those files 197 00:06:54.960 --> 00:06:58.260 off of that solid-state device and free up some extra space. 198 00:06:58.260 --> 00:07:00.690 This is because of the way that a solid-state device works 199 00:07:00.690 --> 00:07:02.040 that when it starts filling up 200 00:07:02.040 --> 00:07:03.930 and having almost no free space, 201 00:07:03.930 --> 00:07:05.520 that device will slow down 202 00:07:05.520 --> 00:07:07.380 a lot in terms of its performance. 203 00:07:07.380 --> 00:07:09.390 So I always recommend keeping at least 10% 204 00:07:09.390 --> 00:07:11.640 of your solid-state device empty and free, 205 00:07:11.640 --> 00:07:14.240 and this will make sure your performance stays high.