Homerr

pwm, thanks for the heads up. Hoopster, are you using Red or Red Pro?

July 24th, 2018 report: https://www.backblaze.com/blog/hard-drive-stats-for-q2-2018/ I just read it hoping to figure out which drives to get for my server, I'd like to change out my 4tb parity drives for 8tb ones. I realized their drives are all 7200rpm, I'd still consider slower rpm drives for low heat and noise. The data suggests Seagate ST8000DM002 ($350, looks like they are phasing out) and ST8000NM0055 ($260) drives have low fail rates. The 0055 drive has low Newegg reviews (taken with a grain of salt) and Amazon reviews seem to indicate they are OEM and could be return/recertified drives. And the last 4 Seagates that I have bought have all died within a year, so I'm carrying a bit of a chip on my shoulder about them. The HGST HUH728080ALE600 ($280) is another contender in the 7200rpm class with better reviews and a few DOA duds noted. Anyone have input on other drives to consider which might be slower, like NAS, Purple, or the like?

I'm in need of some help to make sure I don't screw things up moving to your Crashplan Pro app. I'm a complete noob with linux so I struggle to understand some of the basics of this discussion, my confidence is feeling pretty low at the moment. I'm decently competent at Windows stuff, but I learned just enough about unRAID to get it working a couple of years ago, the official setup video was 90% of it. Normally I'd cowboy it ahead - but with this I just want to back up my data, not really learn a bunch of linux just so I can install this app. And I'd like to avoid reuploading 3.2tb of backup, if possible. Here's where I'm at currently: Have been using Crashplan Home and gfjardim app. Upgraded to Crashplan Pro around October. Had a couple of drives fail, been getting back in to relearning how I set unRAID up , have that resolved. Recently I restarted the rashplan/noVNC thing and now I get a black webUI box instead of Crashplan Pro as before. I've not done anything or attempted any install of this app. 1. Have I borked something by restarting the gfjardim app? 2. My head has exploded just trying to understand the Djoss/jlesage GitHub 'Quick Start' guide. Does the screenshot of the gfjardim app help out? Can you walk me through tailoring this command for my setup? 3. I'm looking at the Github doc about 'Taking Over Existing Backup', has anyone done a video of this sequence? 4. Am I not enough of a linux user to use this? Should I consider starting my backup over with something else like Cloudberry + Backblaze B2 which I've read a couple of users referring to? Another can of worms, but if it is more 'user friendly' I would strongly consider it. It costs a little more, but not the end of the world. I'm humbled by the level of discussion and thank you for any input.

It looks as though it rebuilt successfully with no data loss. Thanks for all the help.

I've been sick a few days and I'm just getting back to this. I did get two WD 3tb drives in the meantime and they are available and just out of the box. I tried FSProxy, didn't seem to want to mount the drive. Ubuntu/VirtualBox I got lost. Knoppix worked and the file manager could see the 2tb Samsung in question, but didn't want to mount it either (it would mount the C: drive of the PC I was using just fine)...something about checking the log file...again I was a bit lost in this OS. Does any of this change the approach to take? unRAID is sitting at a stopped array state currently. p.s. The SMART on the Samsung is all green except a warning ! on "UltraDMA CRC Errors".

Thanks Squid and Gary for the tips. Yes, I do have a parity disk. Just haven't run a parity-check since all data moved on (I ran one about a week ago with only 3-4 disks in the system then). I'll get another disk tomorrow and try Gary's #1. I'd still like to verify data per #2. I already hooked it up to my Win10 PC, I could see it in under Disk Management. I didn't do anything, couldn't see it in Explorer (wasn't sure if I would). It's been 7-8 years since I've booted up a Linux distro off a DVD, any one in particular I should seek out?

I'm new to using unRAID. I've been building my array the last week or so and just got all my data on the array (Parity+11 drives+cache). I haven't run a full parity check yet, and there is still data on the Cache disk to move in to the array with Mover. I noticed on of my drive bays was what looked like at an odd angle (in a Norco 20 bay case) so I pressed on it and unRAID immediately popped a red X on Disk 8 with "Device is disabled, contents emulated" when I hover over the X. I tried powering down, reseated the drive/cage, powered back up, started the array, and still it shows a red X. Disk 8 is nearly maxed with data. What should I do?

I want to stop the Mover so that I can stop the array to add another drive. I don't have oodles of spare drive capacity and I'm still transferring data in to the array off one drive through another PC and then adding that drive in to the array. Mover wants to do it's job and wants to move data, but it can take a few hours to over a day to do this. Just makes this chaos of moving to this new OS a bit stressful and drawn out. With a 'Stop Mover' button I could manage this transfer on my schedule instead of Mover's schedule. I have some shares I was transferring in that had several hundred thousand of small files and other shares with large video files. Mover was running terribly slow on the small file ones and I was figuring out the transfer rates where it was going to take up to 27 days to run everything through the Cache/Mover. I posted elsewhere here and it was suggested I disable Cache for those shares with all the small files, I did this and that helped to bypass this issue. Even with Cache/Mover dealing with only the larger video files it still can take hours and hours that the completion of does not necessarily align with not sleeping or being at work so I can lose near 8 hours of transfer time when I need to intervene. I'm getting down to 2 more disks to add and I'm in a pinch of space I hoped not to be in. A new disk will be answer to give me extra room that I need. Honestly, knowing what I know now - I just wouldn't have even bothered with installing a cache disk for this initial setup. I would have waited until all my data was transferred in (one drive inevitably is the 'slop' in the leap frog of getting data off drives, which could have been my drive I'm using as Cache drive). Which kind of defeats some of the use of the cache in a way. Once this initial phase is over I don't see it being a problem, but during this I feel like I keep bumping against Mover. It's part of my learning curve. Edit: Yes, Mover did not stop and moved everything off the Cache by this morning.

Just an update, 'pkill mover' is nearing 3 hours and over 220gb moved and not stopped Mover yet! This is all with larger 700mb to 10gb video files.

Thanks for the follow up!

Yes, I tried 'killall rsync', that's from the thread I first referenced. Nothing happened. Seemed like it should have work. I did search further and try another command promt input (can't find it atm) that seemed to just turn off the mover in the scheduler. But scheduler did run until apparently it could fit no more on the array drives, each drive had less than 5mb left. Still, I think a 'Stop Mover' button would be a great addition. I know I was a bit stressed over this and how I was going to solve the issue while the Mover was running.

Well, after getting around 70 messages all day about how my array was full I got home and Mover had stopped finally. I was able to stop the array and add disks.

I've been searching and searching and have found a few commands to type in the terminal to no effect. But I'm having this problem that mrjofus1959 posted in 2011 and there still doesn't seem to be any real implementation to the issue. I am stuck in the loop now myself. Can't stop the array to add space because I can't kill Mover. This is VERY frustrating and seems like such a simple thing to add a 'Stop Mover' button. Meanwhile I'm being spammed with red low disk space alerts every few minutes. And I don't really know what I'm doing at the command line so I'm not going further there. As a new user to this product I can't understate how much this feels like a design flaw.