August 11, 20169 yr My new unRAID server components have all shown up except for the ECC memory which should be here early today. I know it would probably be a good idea to run some tests on my hardware to insure its integrity, but I'm not sure where to begin. I have two new Hitachi Deskstar NAS 4TB HDDs; is there some sort of program or command I should run before committing them to the array? As for the memory, is memtest still the go to memory tester? Anything else I should be doing before diving into this?
August 11, 20169 yr My new unRAID server components have all shown up except for the ECC memory which should be here early today. I know it would probably be a good idea to run some tests on my hardware to insure its integrity, but I'm not sure where to begin. I have two new Hitachi Deskstar NAS 4TB HDDs; is there some sort of program or command I should run before committing them to the array? As for the memory, is memtest still the go to memory tester? Anything else I should be doing before diving into this? When building a new server I usually do a burn in test on the drives and the memory only. Memtest is certainly the way to go for the memory. It's included in the unRAID OS and selectable from the boot menu. Leave it running overnight. If you wake up and there are no errors then I'd be satisfied all is well. As for your disks I would say that it is mandatory for you to stress test these before committing your precious data to them. Infant mortality in mechanical HDD's has always been an issue. There isn't a tool built into unRAID to do this for you but there are some excellent community tools which you can use. The most common tool used for this for unRAID Is called the preclear script: http://lime-technology.com/wiki/index.php/Configuration_Tutorial# A utility to "burn-in" a new disk, before adding it to your array has been requested several times. Also requested is a process to "pre-clear" a hard disk before adding it to your array. When a special "signature" is detected, the lengthy "clearing" step otherwise performed by unRAID is skipped. There are two versions of the script. An original created by Joe.L and a "Faster" version created by Bjp999: http://lime-technology.com/forum/index.php?topic=2817.0 http://lime-technology.com/forum/index.php?topic=32564.210 As indicated in the wiki link these scripts are run from the command line, so some familiarity with the CLI is required. A plugin was created by gfjardim to act as a frontend to the preclear scripts: http://lime-technology.com/forum/index.php?topic=39985.0 A further version of the plugin was created (which is in beta currently) which includes a re-written version of the script (inspired by the originals) but I reference this here for completeness sake only. For now I'd recommend using the stable version of the plugin and one of the well used and tested scripts. As is mentioned in the links above unRAID requires that all disks have a "cleared" signature on them before they can be added to the array. As far back as I can remember unRAID has always done this natively when you add a disk (read all, write all, post write read all). The issue was that during this time the Array became unresponsive and your server was not usable. This started to become more of an issue as disks started getting bigger and the process took longer. Therefore the script was also used to address this issue. By being able to add this signature prior to adding the disk to the array meant no downtime. It's important to note however that v6.2 does not have this issue and will clear a disk for you without making the server appear offline. Back to the testing reason for the script we go. As I note about 6.2 I see really no reason for the script other than for preparation / burn testing. Which is what I use it for. My regime to ensure my disks are adequately tested prior to deployment is: SMART short test. 3 cycles preclear. SMART long test. If each of those complete with no issue then I consider the disk ready for deployment. If not I RMA the drive. There are of course other tools that you can use (e.g HD Sentinal or a vendors testing tool). Lots of users do use these. Dock the drive connected to a Windows machine - run one of those tools. Get thier assurance that the drive is ok and then add to unRAID. I would note however that these tools won't add the "cleared" flag (like the preclear script will) to the disk so to use these won't skip that step when adding the disk to the array. Hence why I prefer the preclear script as my tool of choice. If you do use the preclear script and want some help interpreting the reults. Post away. I'm going to stop writing now. EDIT: apparently there have been some issues with preclearing disks on v6.2. I am not sure which version of unRAID you are going to install so here is a link to read documenting what people have experienced and respective fixes. That post onwards. http://lime-technology.com/forum/index.php?topic=13054.msg463812#msg463812
August 11, 20169 yr Author Wow, danioj thank you so much. I have read a bit about the preclear utility previously and that sounds like the way to go. It's also extremely convenient for both preclear and memtest to be included with unRAID. I'm planning on using 6.19 until 6.2 is stable. Thanks again, I think you've given me all the information I need!
August 11, 20169 yr Pre-clear isn't included in the unRAID download. You need to add it. As danioj said, it started life as a script and later a plugin wrapper was written for it. You'll need to install them separately. You can get the plugin from here: http://lime-technology.com/forum/index.php?topic=39985.0 in the first post of the (very long) thread. Instructions for installing the faster version of the script are here: https://lime-technology.com/forum/index.php?topic=50930.0 This combination works well with unRAID 6.1.9 but when you upgrade to 6.2.0 read this: http://lime-technology.com/forum/index.php?topic=39985.msg477858#msg477858
August 11, 20169 yr Community Expert I always preclear disks for testing purposes also. That being said, it is not a requirement for any disks to be clear until parity has been built, so you can create a new array without clearing any disks if you would rather test the disks some other way on another system.
August 11, 20169 yr As is mentioned in the links above unRAID requires that all disks have a "cleared" signature on them before they can be added to the array. As far back as I can remember unRAID has always done this natively when you add a disk (read all, write all, post write read all). The issue was that during this time the Array became unresponsive and your server was not usable.Small clarification. Unraid has never natively done the read, write, read that preclear does. It's always (and still) only done a write of zeros. It relies on the drive to announce if it has an error while doing that, it doesn't go back and check that the zero was written correctly and can be read without error. That's why thoroughly testing drives with 3rd party tools is still important, even though clearing is now done with the array available.
August 11, 20169 yr As is mentioned in the links above unRAID requires that all disks have a "cleared" signature on them before they can be added to the array. As far back as I can remember unRAID has always done this natively when you add a disk (read all, write all, post write read all). The issue was that during this time the Array became unresponsive and your server was not usable.Small clarification. Unraid has never natively done the read, write, read that preclear does. It's always (and still) only done a write of zeros. It relies on the drive to announce if it has an error while doing that, it doesn't go back and check that the zero was written correctly and can be read without error. That's why thoroughly testing drives with 3rd party tools is still important, even though clearing is now done with the array available. I always preclear disks for testing purposes also. That being said, it is not a requirement for any disks to be clear until parity has been built, so you can create a new array without clearing any disks if you would rather test the disks some other way on another system. I love this forum. That's two bits of information I never knew. Excellent Clarifications
August 11, 20169 yr Author Thanks for the responses everyone. I have everything assembled and memtest is currently running. 4 passes so far and no errors. I'm gonna let it run overnight and then start on the disks tomorrow.
August 17, 20169 yr Author Preclear just finished on both of my disks. 3 passes total. Just want to verify that there were no problems before I add them to an array. sdb ============================================================================ ** Changed attributes in files: /tmp/smart_start_sdb /tmp/smart_finish_sdb ATTRIBUTE NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS RAW_VALUE Temperature_Celsius = 162 206 0 ok 37 No SMART attributes are FAILING_NOW 0 sectors were pending re-allocation before the start of the preclear. 0 sectors were pending re-allocation after pre-read in cycle 1 of 3. 0 sectors were pending re-allocation after zero of disk in cycle 1 of 3. 0 sectors were pending re-allocation after post-read in cycle 1 of 3. 0 sectors were pending re-allocation after zero of disk in cycle 2 of 3. 0 sectors were pending re-allocation after post-read in cycle 2 of 3. 0 sectors were pending re-allocation after zero of disk in cycle 3 of 3. 0 sectors are pending re-allocation at the end of the preclear, the number of sectors pending re-allocation did not change. 0 sectors had been re-allocated before the start of the preclear. 0 sectors are re-allocated at the end of the preclear, the number of sectors re-allocated did not change. root@unRAID:/usr/local/emhttp# sdc ========================================================================1.15b == HGSTHDN724040ALE640 PK1334PEJNRD1S == Disk /dev/sdc has been successfully precleared == with a starting sector of 1 ============================================================================ ** Changed attributes in files: /tmp/smart_start_sdc /tmp/smart_finish_sdc ATTRIBUTE NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS RAW_VALUE Temperature_Celsius = 162 206 0 ok 37 No SMART attributes are FAILING_NOW 0 sectors were pending re-allocation before the start of the preclear. 0 sectors were pending re-allocation after pre-read in cycle 1 of 3. 0 sectors were pending re-allocation after zero of disk in cycle 1 of 3. 0 sectors were pending re-allocation after post-read in cycle 1 of 3. 0 sectors were pending re-allocation after zero of disk in cycle 2 of 3. 0 sectors were pending re-allocation after post-read in cycle 2 of 3. 0 sectors were pending re-allocation after zero of disk in cycle 3 of 3. 0 sectors are pending re-allocation at the end of the preclear, the number of sectors pending re-allocation did not change. 0 sectors had been re-allocated before the start of the preclear. 0 sectors are re-allocated at the end of the preclear, the number of sectors re-allocated did not change. root@unRAID:/usr/local/emhttp# Thanks again everyone for the help, this forum is great.
August 17, 20169 yr Author Both disks are looking good. Enjoy! Thanks John_M. I started the array, but now the webui seems to have timed out or something. I just get the "spinning circle" loading indicator on the tab and at the bottom left of the webui it says "Spinning up all drives..." Is this normal? Refreshing the page and trying to open up the webui in a new tab or browser isn't working either. EDIT: It became responsive again, but the array was still stopped. Nothing seems to have happened. Trying it again now.
August 17, 20169 yr No, it shouldn't get stuck there. Hopefully a reboot will get things going again but, before you do, if you have a keyboard connected type diagnostics which will put a zip file in the logs folder of your usb flash device. Attach it to your next post. root@Lapulapu:~# diagnostics Starting diagnostics collection... done. ZIP file '/boot/logs/lapulapu-diagnostics-20160818-0006.zip' created. root@Lapulapu:~#
August 17, 20169 yr Author No, it shouldn't get stuck there. Hopefully a reboot will get things going again but, before you do, if you have a keyboard connected type diagnostics which will put a zip file in the logs folder of your usb flash device. Attach it to your next post. root@Lapulapu:~# diagnostics Starting diagnostics collection... done. ZIP file '/boot/logs/lapulapu-diagnostics-20160818-0006.zip' created. root@Lapulapu:~# Thanks John, rebooting seems to have solved the issue. I created a diagnostics file anyways and I will upload when the parity sync is finished.
August 18, 20169 yr Author I've gotten this error twice now during parity sync: array health report [FAIL] It doesn't seem to effect anything, and I've checked the SMART attributes on both drives and they both look find. Is it just saying this because there is no parity yet?
Archived
This topic is now archived and is closed to further replies.