emo_boy

It finished rebuilding with no sync errors. So if I get another "Parity disk in error state" warning in the next couple of days, do you think I need to start replacing wires?

Sounds good. I now stopped the array, unassigned the parity drive, started the array, stopped and re-assigned the parity drive, and started it back up for a parity sync rebuild. Crossing my fingers that fixed it.

Okay, I did a clean power down. Secured all the connections and then re-booted. Attached is the new diagnostics log. Note: while doing this I got the same looking main page (with a red X on my parity drive). I'm not sure a full parity check occurred. Would you recommend a stopped the array, unassigned the disk, started the array, stopped the array, then assigned the disk, and do a full re-build? tower-diagnostics-20190206-2149.zip

Seven days ago I received notice that my Parity disk is in error state. After some research, I stopped the array, unassigned the disk, started the array, stopped the array, then assigned the disk and started it back up and rebuild the parity drive. The parity drive re-built with no errors and the whole server was working just fine up until today, 7 days later. I've attached a screenshot of the main page and the diagnostics log. Any thoughts on what to do next? Could it be just a loose wire? Or is the disk faulty (hopefully not the case since it's my newest drive, only about 1.5 years old)? tower-diagnostics-20190206-2125.zip

Yeah, I was wondering if it was the power supply since now some drives (such as the Parity) are starting to not show up after the first reboot that aren't on the controller card. Unfortunately, I don't have a spare power supply at the moment, so I'll need to decide which to buy and try first (a new HBA or power supply).

It hasn't always been this way. I've ran Unraid with this machine for 5 years. But it's been like this for a while, I'd say for the past 6 months to a year it's been acting up after a reboot. It almost always takes 2-5 reboots before the drives all come back properly for Unraid.

So I began to do some research on different HBA's, given your suggestion of updating my current SUPERMICRO AOC-SASLP-MV8 to an LSI (I'm considering an LSI SAS9211-8I). Anyways, I wanted to go through some troubleshooting on the MV8 first. So I plugged it into a different PCI-e slot (made sure it was snug), used the other port on the controller card (it has 2) and disabled int13. However, when I booted back up, same symptoms occurred (my parity drive and a drive plugged into the controller card) weren't detected nor assigned when Unraid started. I made sure to get a diagnostics report, see attached. When I rebooted the server once more, everything worked just fine. Not sure what the issue is. Typically, once it finally boots up clean everything runs fine until I need to reboot again. But it's definitely annoying to have to reboot 5+ times after every update, etc. Any thoughts? tower-diagnostics-20180313-1747.zip

Solved. I checked the connections and restarted, everything is good after the parity check. I should have thought to just check the connections first. It seems like every time I boot up, one of the four drives (it's not always the same one) on the controller SATA card I have isn't found when Unraid starts. So I always have to power down, and double check all the connections and boot back up. After 2 or 3 tries, the drive shows up and everything works fine. But I've never had a problem like this, with a drive off the controller and causing an issue that's not at first boot. I guess it may be time to look at upgrading some of the equipment like the motherboard.

I received an email saying the "Parity disk in error state" and now have the red X next to my parity drive. The SMART diagnosis on that drive says, "A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options, unraid". Not sure how to proceed, any help would be great. I'm running 6.4.1 and I've attached my Syslog here. tower-diagnostics-20180308-0647.zip

Thanks Frank. I went ahead and used a different Flash drive, loaded 6.1.9 on it, with my Shares folder and .key and went through the replacement key process. I assigned the drives back up and did a sync on it. So now everything is back working again with 6.1.9. I may just wait a bit on trying to upgrade to 6.2. I'm sure it was mostly due to the Flash drive, but I've read a couple other threads having issues booting up and getting stuck on /bzroot during the loading process. Might wait to see if Limetech can replicate and/or fix this potential issue.

Been running 6.1.9 smoothly since it came out, stock with no plugins. I attempted to upgrade to 6.2 via the webGUI today, but when I stopped my array and then rebooted the server like it prompted me to, I couldn't get past the initial boot unRaid selection screen (I saw a new screen that contained the GUI boot option). No matter which option I tried (normal unRaid boot, unRaid safe boot without plugins), it would get stuck on loading /bzroot. It would just sit there and not move past that option. I ejected the flash drive and plugged it into my Windows machine. I started to notice some issues with the Flash drive, Windows kept saying the device may not be working properly. I copied all the files onto my desktop, but the files in the "previous" folder (I'm assuming from the update via the webGUI) appear to have nothing in them (eg changes.txt says it's 18 KB but when I try to open it in Notepad, no text is displayed). So after I copied the files off the Flash drive and noticed some weirdness, I decided to try installing 6.2 fresh the "old-school" way. I copied/overwrote files from the 6.2 download .zip. But then I noticed some files (namely files and sub directories under the "syslinux" folder) had flunky characters in their filename. It seems like I can manually create the folders on the Flash drive just fine. But copying them over from the .zip folder always ends up with corrupt files in that folder. See attached screenshot. So it appears that I'm stuck, I can't seem to get the Flash drive to work correctly. I've tried to re-format the Flash drive and copy all the 6.2 files over with no avail. It says something to the effect of"can't find config file" when I boot. I have an older backup copy of the flash drive (a 6.1.7 from earlier this year, which I've added a drive to since), but when I format and use those files and boot it the screen just says "SYSLINUX 6.03 EDD". Any thoughts on what to do? Think it's a bad Flash drive? I have an older (but hopefully better functioning) Flash drive that I could use right away.

Okay, I ran 10 passes of the Memtest86+ and it found 0 errors. So maybe it's not the RAM. Any other suggestions to try? It does appear to be running smooth now. A parity check found 0 errors.

Thanks for the tips guys. Here's what I did. I deleted the folders /config/plugins/dynamix and /config/plugins/dynamix.kvm.manger and the web interface immediately turned back to normal. Then I did a reboot on the server, and everything looked correct again (no jumbled up/corrupt looking text). I'm going to run a parity check (it did say the last check found 42 errors!) and continue to monitor things a little closer to make sure it isn't bad RAM. Since my issue only started right after I did the update, I'm wondering if something got corrupt during the update on the Dynamix side of things. For now, I'm going to mark this as resolved and cross my fingers it's nothing major.

Could it also be corruption of the Flash drive file system? Console corruption makes me think RAM. I guess maybe if the console fonts are reloaded from the flash at some point in the boot process? Temporarily booting just the USB stick in another box and looking for console corruption would be a good test. Here's a dumb question, should I have a plugin for dynamix in my Plugins folder? Not sure if that was brought over from my 5.0.x upgrade, but from what I've read Dynamix webGUI is now built into 6.0. The errors shown on the web interface point to "/boot/config/plugins/dynamix/dynamix.cfg" and other Dynamix issues. Maybe it's just a display issue from that? Not sure what the file is suppose to look like, but I attached the corrupt text looking .cfg file. Or do you think it's still a possible RAM issue? dynamix.cfg

I ran an update from 6.1.6 to 6.1.7 from the web interface. Everything seemed to go fine, when I rebooted the server, it went through it's normal boot process. But at some point along the way, the text starts to glitch out. I attached a screenshot. Furthermore, the web interface doesn't look right either. I've rebooted the server and the same thing happens. The data appears to be fine. Any ideas? Do I dare manually upgrade to 6.1.8 (if I can remember how)?