WEC
Members
-
Joined
-
Last visited
-
Installed this plugin recently and really like the insight it provides at a glance, nice work! The only issue I've observed is that since installing the plugin my zfs pool arrays are no longer spinning down after the timeout. Is there a chance the plugin calling /proc/diskstats is keeping the disks awake? If that is what is happening, maybe a change so that the daemon is only running while there is an active gui session?
-
Just to update this thread, the drive continued to flag issues and was replaced out of caution.
-
I've noticed an issue with rc2 which I'm not able to reproduce or prove easily but I'm providing this report in the hope that it might point the developers to the general area of interest. I have a server with three ZFS pools and I've noticed since updating to rc2 from rc1 that one of my pools named ZFS has a random drive each night drops into a faulted state. For example: After rebooting the drive resilvers and operates as normal and is in good shape SMART wise. Initially, I thought I just had a bad drive and planned to replace it but then I noticed that each night the failing drive was different. Over 5-6 nights I noted that nearly all the drives in the pool had at some point been marked as faulted with no pattern. Different brands of drives, different ages. I moved all the drives over to a different backplane slot to rule out a connection issue. The next thing I noticed is that the issue starts around 4:30-4:40am each day. If you check out the attached syslog which is a notepad I've been keeping each morning of the issues, the drive of choice that morning always develops the fault around the same time. I've checked all my apps and plugins and nothing is running at that time from what I can see. Long story short, I've downgraded to RC1 and the issue hasn't happened since. I hope this report helps in some way but I won't be offended if you need more or can't trace the issue. ecclesserver1-diagnostics-20241231-0932.zip ecclesserver1 syslog.txt
-
So after another preclear with no concerning values I'm considering adding the disk back into my next vdev expansion
-
So a couple of days ago I noticed a drive in my ZFS pool was acting up. zpool status was showing 100-200 write and read errors shortly after trying to move some TV shows to a share that exists on this pool. I shutdown the server, removed the drive caddy from the hot-swappable drive bay, and reinstalled making sure that it was firmly pushed and locked back into the enclosure. After a couple hours I had the same issue with read/write errors in zpool status causing it to be kicked from the pool and the pool marked as degraded. There were a bunch of IO errors in the logs and the reported pending sector, CRC and offline uncorrectable jumped up to around 32. I was convinced this nearly 4 year old drive was failing so I ordered a replacement. I even recorded the swapping out process moving a new drive into the existing shelf location and putting the old drive in a different slot. After resilvering the new drive into the pool and ensuring everything was good I used pre clear to erase the drive and to my surprise the smart appeared to improve as per below: #################################################################################################### # Unraid Server Erase of disk WRD0BJMX # # Cycle 1 of 1. # # # # Step 1 of 1 - Erasing the disk: [11:23:47 @ 195 MB/s] SUCCESS # # # # # # # # # # # # # # # #################################################################################################### # Cycle elapsed time: 11:23:49 | Total elapsed time: 11:23:50 # #################################################################################################### #################################################################################################### # S.M.A.R.T. Status (device type: default) # # # # ATTRIBUTE INITIAL CYCLE 1 STATUS # # Reallocated_Sector_Ct 408 408 - # # Power_On_Hours 33941 33952 Up 11 # # Reported_Uncorrect 1444 1446 Up 2 # # Airflow_Temperature_Cel 15 24 Up 9 ->Failed in Past<- # # Current_Pending_Sector 32 0 Down 32 # # Offline_Uncorrectable 32 0 Down 32 # # UDMA_CRC_Error_Count 8 8 - # # # # # #################################################################################################### # Report genereated on: November 18, 2024 at 06:12:08 # #################################################################################################### --> ATTENTION: Please take a look into the SMART report above for drive health issues. --> RESULT: Erase Finished Successfully! The bit that I'm interested in is that these sector issues went down! which I've not seen before. # Current_Pending_Sector 32 0 Down 32 # # Offline_Uncorrectable 32 0 Down 32 I'm now running a 2 cycle preclear using pre and post reads but has anyone seen this before when a drive seems to bounce back into good health? Assuming the pre clear cycles pass without error I'm unsure if I should trust the drive again. What are your thoughts?
