gfjardim

Just a trivia, guys. Tested both bjp999 and my script on a 2TB disk: bjp999: 21h34m49s gfjardim: 19h06m04s Almost two and a half hours saved is a very good mark, IMO.

Diagnostics attached. Lots of sense.key nonsense. Unsure what matters in this syslog. Perhaps some hardware errors will be noticed by someone else. (@turl or @robj) (FYI this is a Xeon 1220 Sandybridge on a Supermicro X9SCM with 20GB Ram and the affected devices are all connected to an IBM1015 via a Norco 4224 chassis.) I've updated the script to output some debug info. Could you please update and run another instance of Clear on /dev/sdp?

I may have already done that. Not sure if this is what you mean?? 2 days ago I set all 3 disks to do a complete preclear again. The disk that did the full preclear initially did another full preclear including a post read verify as JoeL's script would have done. Disk 2 replicated the previous preclear as follows Could you please look into your syslog for Out Of Memory occurrences?

Yep, figured that already. Could you start a Varify All the Disk and tell if it failed again?

Please hit "Help" and post a copy of your log.

It's safe to update.

Could you please update and test if this error has been corrected?

No unassigned disks at this time. So UD is showing no disks, right?

Same here. I had 3 drivers showing up via unassigned devices as expected. But 3 more array devices show up in the preclear plugin screen, one of which was parity 2. Areca is involved in this server as well using the dynamix SCSI plugin. via Tapatalk Unassigned Devices is working ok for you, guys?

Use this link: https://raw.githubusercontent.com/gfjardim/unRAID-plugins/master/plugins/preclear.disk.plg

bjp999 script has the "Fast post-read verify" option that relies on an executable file (named readvz and located on your flash drive) to verify if the disk is effectively zeroed. My script relies on a completely different approach that make both pre-read and post-read almost as fast as the disk can go. There's no switch, you have the optimal performance out-of-the-box.

Should be replaced by readlink, which is part of Coreutils and installed virtually on every linux machine out there.

Which distro?

It depends. The AOC-SASLP-MV8 has a 600MB/s cap, so if you plan to do a parity check with more than 5 disks connected to it you could have a bottleneck.

I only (very) recently put 6.2 beta on my server. I did not have any issue pre-clearing the second parity disk I have just added to my array. The fix will need to wait until I add/replace one of the existing disks with a larger one. (Otherwise, I have no way to test the process. ) Whatever the fix might be, it must be backwards compatible with the older releases of unRAID. In the interim, you can type this command to "patch" the preclear_disk.sh command First change directory to the directory holding the preclear_disk.sh command. For most, it will be cd /boot then type (or copy from here and paste) the following: sed -i -e "s/print \$9 /print \$8 /" -e "s/sfdisk -R /blockdev --rereadpt /" preclear_disk.sh Your preclear disk script will be edited and should work with the two changes you mentioned. (actually, each occurs in two places, so there are a total of 4 lines changed) Joe L. There's one more fix to be made. Because of dd output change, pre-read and post-read status are showing time instead of speed. I propose a more elaborate command: sed -i -e "s/print \$9 /print \$8 /" -e "s/sfdisk -R /blockdev --rereadpt /" -e "s/ sed -n 3p [^\}]*}'/awk -F',' 'END{print \$NF}'/g" preclear_disk.sh

I'm sure I'm being more than a little dense, but which script is the "it" you're recommending? My script.

1) the quick preclear script by bjp999 has always being supported. The procedure is the same as Joe L. script. 2) if you have Joe L. or bjp999 script inside /boot/config/plugins/preclear.disk/ directory, then you can choose it from the dialog window. Now, because of the recent modifications I made in the code, I believe it's faster than both older scripts.

[glow=red,2,300]UPDATE:[/glow] 2016.10.26 - Add: Copy to Clipboard button on Preview window - Fix: unescaped quotes breaking read of "state.ini" file - Fix: remove PHP "split" function due deprecation - Fix: script - better SMART detection - Fix: script - greatly improve speed and memory footprint in read operations - Fix: script - faster display refresh

I changed the SMART device type detection in the next update. If you could test it then....

Found your issue. Will be fixed in the next update. This is a bug introduced by PHP 7 dropping "split" function. Fixed. A) Those OOM errors can't be easily caught, but I just changed loads of code and the memory footprint has lowered more than 100x in my testings; B) This is because the page updates so fast it's nearly impossible to select anything. I'll search a way to make this possible;

For those with black screen issues, does the latest update improved your situation?

Not true. A small bug was introduced in 2016.9.26b which prevents that version to update. Please consider reinstalling the plugin to reenable the update detection.

Guys, I'm back and will start to work soon. Enviado de meu SM-N910C usando Tapatalk

Appreciate you stopping by! I did that, but it's still showing results, and the X says it will clear stats, but doesn't. I guess I could restart it by hand, but... ewww! Any other tips? I rebooted the server, and it let me restart the preclear. Now it's telling me that the preclear is done. gfjardim, you've done some amazing work, brother! That's a 4TB drive cleared in... 12 hours!! OK, probably not. preclear_disk.sh -t /dev/sdh Yields all the instructions on how to call preclear_disk.sh followed by: Sorry: Device /dev/sdh is busy.: 1 All that's in /boot/preclear_reports is a report for the drive that died last week. It's dated Sept 17th, and is full of: tput: unknown terminal "screen" (6536 bytes worth). I redownloaded the original preclear_disk.zip, extracted preclear_disk.sh from it and compared SHA256 fingerprints from the fresh copy v the copy I'm actually using, and they match. I know gfjardim is busy, so I probably won't get a response from him for a couple of days (I get it - real life. Glad he's got one!). Anyone else have any suggestions? No, the drive's not hot swappable, unfortunately. (I guess I need to watch Office Space again, 'cause I totally didn't get the reference...) I'm attaching screenshots of the "preview" (clicking the blue eye), since it won't let me copy/pasta the text. Remove the plugin and reinstall it again.

Guys, I wish I could give proper support right now, but I can't. 1) SSD wear out: if you want to add a SSD to the array, you should run blkdiscard /dev/sdX an then run /usr/local/emhttp/plugins/preclear.disk/script/preclear_disk.sh --preclear-only /dev/sdX on the command line. I'll make this accessible by the GUI when I have time; 2) Killed session of preclear: if the X icon isn't working, you can kill it by typing the following code: tmux kill-session -t preclear_disk_sdX 3) Number of passes: the more the merrier, IMHO. If you have time to let it preclear for a week, then you should do it. Mechanical failures are extremely rare these days, so mechanical wear out is debatable.