Juicebox Posted February 4, 2022 Share Posted February 4, 2022 Hello! I came home yesterday and none of my home automation was working. The Docker that manages the zwave integration on my Unraid box had stopped and was throwing errors that the filesystem was read-only. I thought it was strange, as I certainly hadn't touched anything and it had been running smoothly for months and months. I decided to reboot my Unraid box, but noticed that it was struggling to stop the array (taking a really long time). It eventually just rebooted itself. When it came back up, it automatically started a parity check. There were a couple yellow notification bubbles that there were some parity errors. The attached screenshot shows 1175 errors on the Parity drive (it used to read 0, iirc), but the Parity history shows 0, and it shows 'Parity is valid.' Should I be concerned? Any specific action I should take? Also strangely concerned that one of my Dockers thought it's filesystem was turned to read-only, but I didn't get any logs from that, and it appears to be working fine after the reboot. Thanks in advance for any support anyone can provide. ganymede-diagnostics-20220204-1215.zip Quote Link to comment
trurl Posted February 4, 2022 Share Posted February 4, 2022 The Errors column on Main is showing I/O errors on each of the drives, whether parity or some other. Often I/O errors are connection problems instead of disk problems, but see below. The Parity History and Parity Check numbers are about parity sync errors, which might not necessarily be the parity drive itself, it means parity doesn't agree with the parity calculation from reading all the data disks. So, these numbers are indicating totally different things, though I/O errors (on any disk) might cause parity check to have sync errors. SMART for parity disk does show a few Reported Uncorrect. You should see a SMART warning for that disk on the Dashboard page, and if you click on the disk and go to its Attributes you will see that highlighted. Do any of your other disks show warnings like that? You should setup Notifications to alert you immediately by email or other agent as soon as a problem (such as disk health) is detected. Parity hasn't had any SMART test run on it. Click on the disk to get to its page, disable spindown, and start a SMART extended self-test. Unrelated, but your appdata and system shares have files on the array. Quote Link to comment
Juicebox Posted February 5, 2022 Author Share Posted February 5, 2022 (edited) Thank you -- that all makes sense! I attached my SMART report... (didn't realize it took that long to do a SMART check!) With my limited understanding of it, I'm assuming my parity drive is about to crap out and replacement is wise? Once I pop a new drive in, is best practice to just stop the array, assign new drive as parity, start the array and just let it rebuild? None of my other disks show any warnings like that. On the unrelated note -- I have appdata set to Only : Cache, and system is set to Prefer : Cache. If I update system to Only : Cache, what else would I need to do? Why would appdata have files on the array if it's set to Only : Cache? Thanks again for your help! Edit: Upon googling, it sounds like Only : Cache is wrong, and it should be set to Prefer : Cache? I set appdata to Prefer, stopped all my Docker containers, and invoked the mover. Hope it really is that easy. ganymede-smart-20220205-0820.zip Edited February 5, 2022 by Juicebox Additional detail Quote Link to comment
itimpi Posted February 5, 2022 Share Posted February 5, 2022 5 minutes ago, Juicebox said: Why would appdata have files on the array if it's set to Only : Cache? because the mover application ignores shares that have the Only setting. It also ignores open files so you typically have to (temporarily) disable the Docker/VM services when trying to get everything on the cache. 7 minutes ago, Juicebox said: (didn't realize it took that long to do a SMART check!) If it was the Extended test then typically around 2 hours per TB 9 minutes ago, Juicebox said: Once I pop a new drive in, is best practice to just stop the array, assign new drive as parity, start the array and just let it rebuild? Yes Quote Link to comment
trurl Posted February 5, 2022 Share Posted February 5, 2022 3 hours ago, Juicebox said: replacement is wise? Yes # 1 Extended offline Completed: read failure 10% 28061 7748072120 Quote Link to comment
Juicebox Posted February 6, 2022 Author Share Posted February 6, 2022 Replacement drive on the way! Thank you trurl and itimpi! Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.