<div dir="ltr"><div><div dir="ltr">On Fri, Jan 28, 2022 at 12:39 PM Khalid Baheyeldin <<a href="mailto:kb@2bits.com">kb@2bits.com</a>> wrote:<br></div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div dir="ltr">On Sun, Jan 23, 2022 at 6:57 PM Ron Singh <<a href="mailto:ronsingh149@gmail.com" target="_blank">ronsingh149@gmail.com</a>> wrote:<br></div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div style="font-family:tahoma,sans-serif">Doing a backup of 13.5 Gig of data(mostly small files) from my MS Onedrive(locally sync'd on my SSD) onto a 5400RPM 512G USB3 2.5" HDD(Hitachi), I am seeing Thunar sucking up 18%-26% of my i7-2640M CPU for the duration. PC is running Mint 20.2 Xfce.<br></div></div></blockquote><div><br></div><div>I got a USB 3.0 dock, and connected it to the same 'server' (i.e. desktop) that had the now dead e-SATA</div><div>dock. The server has only USB 2.0 ports and will be replaced soon with another desktop that has USB 3.0.<br></div><div><br></div><div>Here is a detailed capture of performance data before, during and after the backup at 5 minute intervals.</div><div>The tool is dstat. <br></div><div><br></div><div>The incremental backup size was 17GB. <br></div><div><br></div><div>As you can see, wait for i/o is high, but that is expected for any intensive disk operation.</div><div>What is also high is user time, and I think this is because of using USB and not SATA.<br></div><div><br></div></div><div class="gmail_quote"><span style="font-family:monospace">stem---- --total-cpu-usage-- ---load-avg--- ---procs--- ---system-- -dsk/total- --io/total- <br>ime |usr sys idl wai stl| 1m 5m 15m |run blk new| int csw | read writ| read writ|<br>03:56:37| 4 0 95 0 0|0.26 0.22 0.19|2.0 0 0.1| 296 294 | 0 42k| 0 3.56 |<br>04:01:37| 4 1 94 2 0|0.73 0.32 0.22|3.0 1.0 0.3| 358 358 |5644k 45k|50.1 4.10 |<br>04:06:37| 7 2 78 13 0|1.92 1.11 0.58|3.0 1.0 0.1|1119 1254 | 31M 1365k| 418 10.0 |<br>04:11:37| 20 3 50 28 0|2.84 2.04 1.12|3.0 0 0.5|2244 3133 | 25M 16M| 336 47.8 |<br>04:16:37| 18 2 50 29 0|3.21 2.66 1.65|1.0 2.0 0.2|1877 2470 | 19M 14M| 245 46.2 |<br>04:21:37| 21 3 42 34 0|2.59 2.72 1.97|2.0 2.0 0.2|2030 2752 | 22M 22M| 226 57.1 |<br>04:26:37| 7 1 71 20 0|1.30 1.96 1.87| 0 2.0 0.0|1081 1381 | 10M 4037k| 227 13.3 |<br>04:31:37| 6 1 84 9 0|0.43 1.24 1.60|5.0 0 0.3| 740 979 |5083k 933k| 113 7.02 |<br>04:36:37| 4 0 95 0 0|0.29 0.66 1.24|1.0 0 0.0| 314 318 |5338B 43k|0.10 3.72 |</span><br></div><div class="gmail_quote"><br></div><div class="gmail_quote">When I swap the servers, I will check if USB 3.0 is faster, and uses less CPU time. <br></div></div>
</blockquote></div><br clear="all"></div><div>I swapped the server with a newer (i.e. 2013) motherboard that has USB3.0.</div><div><br></div><div>As you can see, CPU usage is still very high, and even higher than USB 2.0. <br></div><div><span style="font-family:monospace"><br><span style="font-family:monospace">stem---- --total-cpu-usage-- ---load-avg--- ---procs--- ---system-- -dsk/total- --io/total- <br>ime |usr sys idl wai stl| 1m 5m 15m |run blk new| int csw | read writ| read writ<br></span>03:42:26| 2 0 97 0 0|0.19 0.20 0.18|3.0 0 0.5| 290 381 | 0 42k| 0 3.57<br>03:47:26| 2 0 97 0 0|0.15 0.15 0.16|4.0 0 0.1| 308 398 | 0 40k| 0 3.19<br>03:52:26| 2 0 97 0 0|0.15 0.14 0.15|2.0 0 0.5| 281 370 | 0 41k| 0 3.41<br>03:57:26| 2 0 97 0 0|0.16 0.13 0.14|1.0 0 0.1| 284 371 | 0 40k| 0 3.28<br>04:02:26| 13 2 83 2 0|1.85 0.65 0.31|5.0 0 0.3|1800 2697 | 21M 6950k| 206 18.1<br>04:07:26| 58 5 29 7 0|3.42 2.32 1.13|1.0 2.0 0.1|5469 7814 | 57M 53M| 515 120<br>04:12:26| 50 5 37 7 0|2.42 2.58 1.59|4.0 0 0.5|5593 8179 | 58M 36M| 601 88.5<br>04:17:26| 61 6 25 8 0|3.30 3.15 2.13|5.0 0 0.1|6013 8730 | 65M 60M| 547 148<br>04:22:26| 58 6 28 8 0|3.15 3.23 2.47|3.0 1.0 0.2|5830 8408 | 61M 56M| 561 141<br>04:27:26| 55 5 34 6 0|2.84 2.96 2.56|4.0 0 0.1|5636 8276 | 57M 31M| 555 80.5<br>04:32:26| 55 5 32 8 0|3.18 3.05 2.71|1.0 1.0 0.2|5641 8264 | 61M 42M| 544 103<br>04:37:26| 50 5 36 10 0|3.50 3.13 2.83| 0 2.0 0.1|5389 7839 | 55M 39M| 549 96.2<br>04:42:26| 21 3 71 5 0|0.37 1.74 2.35| 0 0 0.6|2593 3842 | 25M 14M| 247 37.6<br>04:47:26| 2 0 97 0 0|0.15 0.70 1.72| 0 0 0.1| 298 387 | 16k 40k|0.50 3.17<br>04:52:26| 2 0 97 0 0|0.13 0.31 1.26|1.0 0 0.5| 297 387 |1830B 39k|0.06 3.37<br>04:57:26| 2 0 97 0 0|0.13 0.18 0.93|1.0 0 0.0| 284 370 | 68B 41k|0.02 3.42</span><br></div><div><br></div><div>All this reminds me why I opted for e-SATA many years ago. <br></div><div>-- <br><div dir="ltr" class="gmail_signature">Khalid M. Baheyeldin<br><a href="http://2bits.com" target="_blank">2bits.com</a>, Inc.<br>Fast Reliable Drupal<br>Drupal performance optimization, hosting and consulting.<br>"Sooner or later, this combustible mixture of ignorance and power is going to blow up in our faces." -- Dr. Carl Sagan</div></div></div>