FQs19

After restarting my server, unraid correctly showed only Disk 4 as unmountable and gave me the option to Format only that disk. I formatted the disk and it appears there was no data corruption on the other disks. Whew... I'm definitely going to request a new motherboard from Asus since I'm still under warranty. Really hope they have a new one to send me. Thanks for everyone's help.

Got it thanks.

I didn't even notice the crashes until trurl mentioned them and I looked in the logs. I saw no problems when using the gui or adding data to my array prior to trying to add the disk. During all of my troubleshooting over the past couple months, I needed Asus to fix a problem in their motherboard's bios. If I disabled the Intel LAN and WiFi controllers, the boot process would go through a couple errors, but eventually boot after 4 mins. Specifically the errors were CPU Code 7 then TEST CMOS. They eventually provided me with a special firmware that stopped those errors from happening. I thought everything was finally fixed, but I started having issues with the onboard 10GbE port. It would sometimes not connect to my switch unless I pulled the SFP+ module then re-inserted it. At this point, I'm guessing this motherboard is actually bad. I'll contact Asus again and see what they say. The motherboard is still under warranty, but the problem is that there aren't any new ones available. I actually can't find any new TRX40 motherboards for sale at a reasonable price. So I basically spent over $2000 for two components that I can no longer use. 😞 So after I restart the server, should I just use the New Config tool to add all of my data disks including the new Slot 4 disk at once or are there other steps I need to take first? Before I shut it down, I wasn't able to remove the new Slot 4 disk and start the array. Thanks for taking a look JorgeB.

I believe that disk is connected to my LSI 9300-8i. I don't want to power down the server and pull the disk to verify just yet. So if it's not connected to my LSI HBA, it's connected to the motherboard. The NVME drives were all original to unraid from prior to all my problems. I formatted them after starting with a clean 6.10.3 boot drive and reconfiguring everything. I will mention that when I started New Config after booting up with 6.10.3 clean boot drive, I left Slot 4 unpopulated, because I knew I was going to add a disk to that slot after preclearing a new disk.

I'm having a very difficult time believing that my luck is this bad, but here I am with yet another issue trying to do something as simple as adding a disk to my array. I originally had 9 array disks in 6.9.2 before I had multiple drive/hardware/bios issues. I removed one disk that failed in slot 4. I left slot 4 empty while I started to reconfigure unraid using a clean boot drive with 6.10.3. I set up my two cache pools, plugins, and configured all my settings, but left parity disks out for now. I figured why rebuild parity until I have everything set correctly including replacing slot 4 disk that had been empty. Here's where things go bad. I stopped my array. Clicked on slot 4. Picked a new precleared disk (sdf). Started the array. Immediately saw this: I scrolled down to then find unraid wanting to format ALL of my data drives: The disk I added is Disk 4. It was precleared and I even verified that the preclear signature was there using the Preclear plugin before adding the disk to my array since some time had passed and I couldn't remember if I precleared that specific disk. My array was working perfectly before I attempted to add the new disk. I even ran the Fix Common Problems after I adjusted all my plugins, docker, and settings to verify there were no issues. I freaked out and immediately stopped the array. I never clicked on the checkbox to format the drives. After stopping the array and trying to unassign the disk I just added, Disk 4, unraid showed an error stating that Disk 4 was missing and the button to Start the array was grayed out because of 'Too many wrong/or missing disks'. I'm really hoping that all I need to do is use the New Config option under Tools, but I'm not sure that's the right move to do at this point. So I'm attaching my Diagnostics and hoping someone could help me recover from my horrendous luck. threadripper19-diagnostics-20220703-2246.zip

I do have another pc with 10GbE. I'll try that. Thank you.

Wasn't aware of that. Thanks. Would using the DiskSpeed docker be okay to use to test for speed or should I just configure the drives then transfer a file to the NVME from another pc? I don't feel like setting up an iPerf test.

Anyone have an idea why both of my Sabrent Rocket 4.0 NVME drives Precleared at such a ridiculously slow speed of 104 MB/S taking over 3hrs to preclear a 1 TB drive? They're PCIE 4.0 drives on a PCIE 4.0 capable motherboard. When I used unraid 6.9.2 and the previous preclear plugin from gfjardim, I was able to get over 3000 MB/S and it finished in minutes. Attached are my diagnostics and a screenshot of the Identity page. threadripper19-diagnostics-20220703-0923.zip

I wanted to test the drives after all the failures. I was able to get the drives to Preclear after I started the array, but the read/write performance is horrible. the max speed I’m getting is only 130MB/sec.

Starting to recover from hardware and software errors after months of being down. I'm having issues with two of my NVME drives, they're both Sabrent Rocket 4.0 1TB drives. I couldn't mount the Sabrent Rocket NVMEs. After clicking on 'Mount', nothing happened. The drives also wouldn't let me delete the original partitions. After restarting, I still couldn't 'Mount' the drives, but was able to delete the partitions. The problem I'm having now is that when I go to Preclear the NVME drives, none of the drive's info is available. There is also no Disk Log information available, but I can see everything under the devices 'Identity'. Anyone have an idea of what is going on? My hardware is in my signature. threadripper19-smart-20220702-1606.zip threadripper19-diagnostics-20220702-1612.zip

Issue resolved. I formatted my USB flash drive and used USB Creator without customizing the install and was able to log in without entering my old password, just leaving the password field blank.

I've tried it twice and both times it required me to use my old password to log in to unraid. I don't want to try another usb drive since that will make my old boot drive unusable for unraid. Do I have to transfer my key file over to the drive to just test if networking works on a clean boot drive?

I've had many problems with my server this year and I found yet another one. I'm hoping that through my recent troubleshooting I've found the original culprit to all of my problems. My issues started with CRC errors, so I switched to a new case without a backplane. During the transfer from one case to another and then parity checks to repair errors, I lost my cache pool. After finally getting that back up, I started having long boot times and occasional network issues. During my recent troubleshooting, I noticed my motherboard post codes hanging on CPU Code7 and Test CMOS. I'm in touch with Asus to try and find what the problem is exactly, but I think my motherboard got messed up with a recent BIOS update I did to help with my previous issues. I've tried to use the second bios and even a different firmware version, but I still have the same issues. Which is why I submitted a ticket with Asus for help. At this point I'm not able to get my networking to work. To try and eliminate my boot drive as a problem, I used the USB Creator to format my original usb boot drive and install the clean version of 6.9.2. I verified that there were no errors on my boot using using EASEUS and chkdsk. I then transferred only my original key file to the config folder. After installing it into my server and booting, I was expecting to use user = root and the password to be blank. However, leaving the password field blank gave me a wrong user name or password error. I had to use my original root user password to get into unraid. Is that right? Shouldn't a clean install of Unraid with just my original key file ask me to make a strong root user password after first logging in? Or does the key file hold your root user credentials?

After installing the update, my server connected immediately to the 'Mothership'. Thanks for the fix.

I'm still seeing 'not connected to mothership' Tried logging out and in several times with no luck.

I would rather have only 10 backups instead of say 180 if I set 'delete backups if they are this many days old' to 180. The problem currently is that if you have your server down for an extended amount of time (over the amount you set to delete if this old) and you lose a drive in the process of bringing the server back online, like I did, you won't have any backups to restore with. I had it to delete after 15 days, because that was keeping only 15 backups on my drives. It took me almost a month to get my server back up and running. I had several problems to deal with and lost the cache pool where the backups were located in the meantime. If there was an option to set 'keep this many backups' and I set it to 10, then I wouldn't have lost any data. I mean really the option to delete after this many days shouldn't exist. There should just be an option to keep this many backups and delete older ones. If you want to keep a specific backup, you can always copy it to another location.

Will do. Thanks

Just lost one of my cache pools that held my appdata and system shares. I was able to get the pool back up and running with help from @JorgeB, but I lost some data. I would love an option to be included with the current Backup/Restore Appdata utility to be able to 'keep this many backups'. Currently there is only an option to 'delete backups if they are this many days old'

Oh nice. Would you mind sharing that script?

Well, I really appreciate you getting my pool back online. I can reconfigure my dockers and plex. Getting Plex back to the state it was in before this mess is going to be a nightmare for sure. Do you have any suggestions on how I can have a backup of my appdata and system shares on my array, but also use the cache pool? I was going to submit a feature request to 'Keep this number of backups' for the Backup/Restore Appdata Utility. I guess I can try and set up a user script to run an rsync to a separate share on my array. Thoughts?

@JorgeB I restored the superblock on both nvme drives: Here's the diagnostics after rebooting and before starting the array: threadripper19-diagnostics-20220509-1128.zip and here's the diagnostics after starting the array: threadripper19-diagnostics-20220509-1140.zip The nvmecache pool mounted and I can see my data, unfortunately the data isn't all there. The music share isn't that big of a deal. I can always re-rip that stuff, but I don't know about the appdata or system shares. Just noticed that I won't be able to restore my appdata because I had 'Delete backups if they are this many days old:' set to 15. I never noticed that there's no option in Backup/Restore Appdata utility to 'Keep' a certain number of backups. Seems really silly to not have an option to keep a certain amount of backups. How will that affect my Plex docker? What about my system share and the dockers in it?

I didn't even think to turnoff the memory overclock. Thanks for the FAQ link to that. Here's the memory I'm using: and according to the FAQ: I should only go as high as DDR4-3200. Wondering if I should just turn off D.O.C.P and let it run at its base speed of 2400 at 1.2V or manually set the memory to 3200 at 1.35V? I turned off Global C-State, but couldn't find anything in my ASUS ROG Zenith II Extreme Alpha bios for the "Power Supply Idle Control". I also turned off ErP Ready in my bios. I'll reboot with memory set to the default of DDR4-2400 and try restoring the superblocks then send you new diagnostics. Thanks

My unraid server has been one problem after another. I just moved it to a new case that doesn't have a SATA backplane because the backplane was causing CRC errors with my disks, but now I lost one of my NVME Cache Pools, called nvmecache, which had both my Dockers and a couple Shares on it, including my APPDATA. No idea when or how I lost just the one cache pool because I had to restart my server several times including updating my motherboard's BIOS. Unfortunately, I set the Shares To 'Only' and don't have a backup of the files. Will this FAQ procedure help me recover my Cache Pool? I almost feel like I should just wipe absolutely everything including all my data and just start again. I'm spending so much time trying to recover files and drives, that it might be easier for me to take the 12 months of ripping my media to a fresh server. I will absolutely turn off Auto-Start array from now on and never use it again. Funny thing is, I've had Auto-Start off since the start of all my issues. So I'm really not sure what happened. FYI: Using 6.9.2 I have two Cache Pools. One called Arraycache and the other called nvmecache. They both use NVME drives. They are supposed to be Raid1. They have identical drives in each cache pool. This has happened to me before, but last time I only lost my appdata share which I was able to restore after recreating the cache pool and restoring my appdata using the Backup/Restore utility. Any help is appreciated. threadripper19-diagnostics-20220508-2158.zip

I finally had time to move my server into a new case. I moved to a Rosewill RSV-L4500U, which doesn't have a SATA backplane/Hot Swap bays. It has three 5 disk bays. I ordered new SATA power cables because my power supply only has four 1x3 ones. I wish they made 1x5 cables, but I got 1x4 from EVGA 4x SATA Cable (Single). I also got extensions to run to the 5th disk in each bay from my fourth SATA power cable. Cable management is difficult without purchasing custom cables (which I'm not going to do). Wish I could use the software for the MSI MPG Series CORELIQUID K360 AIO I have for it, but I'm just using the motherboard's fan controllers. I have 4 disks (including the 2 Parity disks) connected to my motherboard's TRX40 chipset controller. I have the remaining 8 disks connected to my HBA LSI 9300 8i card. I will have 2 disks that will be connected to my motherboard's ASmedia's controller as spares. I'm currently running a Correcting Parity Check. I will do a Non-Correcting Parity Check after it finishes. I'm sure there will be errors since I stopped the Correcting Parity Check before it finished. I haven't seen any errors being reported since switching to the new case. Fingers crossed that it was the old case's backplane causing all my errors. If you guys have any suggestions on what to do after I complete my parity checks, please let me know. Thanks again for the help.

Status update: Just because I had my server down and because I don't have time to move it to a new case, I ran memtest+86 for over two days. It did 7 passes with no errors. So at least I confirmed my memory is good.