atmaledy

So I've got this issue that has happened twice in probably 6 months. My cache array is generally running great except for this one issue. I have only one single drive (not great I know - I plan to get a second one when budget allows). The drive is a 2TB NVME drive. Things will be running fine and then, randomly, the drive goes missing and the array goes offline. The server cannot see the NVME drive. Reboot does not fix it. The only thing that works is shutting off the server, opening it up and reseating the NVME. That brings it back online fine and on boot everything returns to normal. There appears to be also no errors in the logs when this happens. I have sys logging turned on and there's just nothing in the logs about it but that makes sense because the syslog is logging to the cache drive. I have temporarily moved the syslog share to my array so if this happens again I can hopefully see something in the logs. I've had temperature warnings in the past so I am also wondering if that is part of the reason the NVME shuts down. Has this happened to anybody else?

@JorgeB do you have any other recommendations on how to troubleshoot this?

Thanks. I've enabled the syslog server! Side note - I thought that doing a dump of the logs/diagnostics would be just give me the same info (I'm aware the logs delete on reboot and that syslog solves this). Since I pulled logs before reboot, shouldn't the diagnostics I downloaded contain the same info?

Hey all, I cannot for the life of my figure this one out and I'm hoping I can get some assistance from the community. I'm running Unraid as a media server with a bunch of monitoring tools as well. A few weeks ago, my server started randomly doing this thing where docker basically hangs forever and all the containers become unavailable. The WebUI is fine only on the Array control tab. The second I navigate to another tab the whole WebUI starts timing out/hanging. At that point I SSH in and run top but everything appears to be normal. When I run `docker ps` the command hangs. When I `killall docker` then the WebUI comes back and I'm able to download diagnostics, look at everything. Today when this happened, instead of killing all of docker, I killed `cadvisor` (container monitoring prometheus exporter) and that seemed to bring the system back for 10 seconds or so. The problem is the container is set to auto start unless stopped (I've now changed that for next time so I can better troubleshoot). After killing docker I usually do a reboot and everything comes up clean. I've attached logs. I first get notified things services are down at at 10 Jul 2023 at 07:07am PDT tower-diagnostics-20230710-0713.zip If anybody has any ideas I'd be super appreciative! Thank you!

If anybody can help me out here or has any ideas I’d be super appreciative! I’ve got an MSI mortar and it doesn’t have an option to disable fast boot. I’ve got this annoying issue and every time I reboot I have to go into the bios. It’s so frustrating! any help would be appreciated.

I've got this same issue. I'm using an MSI Mortar motherboard that doesn't seem to have a fast boot option. I've tried every combination of boot order to try and get my system to remember to boot off USB. The only fix for me is to actually manually go to the bios and then proceed to boot (save and exit, no changes). That's the only way I can boot into UNRAID. If anybody has any ideas I'd LOVE some help here...