vohk

I believe you were on the right track. I still haven't narrowed down exactly what the issue is but it's looking to be hardware related.

That is a good point. I don't think it's trying a clean shutdown, but it's possible I didn't leave it long enough. I'll make sure to rule that out next time it happens. I hadn't seen that post on c-states! I'll add that to the list. Thanks!

Ah, I forgot. Plugins are: Appdata Backup Community Applications Compose Manager Fix Common Problems NerdTools (none currently installed) rclone unbalanced Unraid Connect User Scripts

My Unraid server periodically goes unresponsive to network connections. When this happens, attempts to reach the panel via browser fail (DNS probe) and ssh attempts report timeout. No services are accessible. The drive activity light continues to blink. I have syslog writing to the cache drive and no unusual events appear to be generated leading up to the event. This persists until the server is rebooted, at which point it starts up normally after an unexpected shutdown (initiates parity check, etc). The frequency of the event seems to be impacted by docker containers but I can't figure out exactly what the trigger is. When I had a pihole container active (DNS+filtering for my LAN), it would occur at least weekly. After removing that container, I'd gone months without a repeat until it happened twice in the past 24h. I have not added any new containers recently. During the event, all other network devices continue to function normally, and the issue has persisted both using an Asus router and more recently an ISP provided one. The Unraid server functions normally so I'm reasonably confident it isn't the overall network configuration. I don't have a spare monitor, so I have not been able to try access the console directly during an event. That's on my list. Anything specific to look for and any other troubleshooting suggestions would be very welcome! (Anonymized diagnostics attached. Syslog for the last year can be provided if helpful but as mentioned, there doesn't appear to be any events of any sort when it occurs) Unraid Version: 6.12.14; all release versions going back at least a year. System specifications: Motherboard: MSI MS-7A31 CPU: AMD 3700x Memory: 32GB DDR4 Storage: 512GB nvme cache drive with appdata and; 2x16TB HDD array Network settings: active-backup IPV4-only, with static IP set. Default Gateway is the router, which also has the same static IP assigned. DNS is static assigned as CF, Google, and the Gateway. box-diagnostics-20241224-1418.zip

That would make sense. The board is getting pretty ancient.

Good thought- I've tried 3, 1x3.0 and 2x2.0 ports- same error, only different usb/port number listed. And I am certain all 3 ports are fully functional, with previous versions of UNRAID and other OSs.

waiting up to 30 sec for device with label UNRAID to come online ... xhci_hcd 0000 :00 :10.1: Error while assigning device slot ID xhci_hdc 0000 :00 :101.: Max number of devices this xHCI host supports is 32. usb usb8-port1: couldn't allocate usb_device not found - press ENTER key to reboot ... After updating UNRAID through the web interface, both with 6.11.0 and 6.11.1, the system fails to boot with the above message. No hardware has been changed and the system was running and booting fine on 6.10.3 previously. I was able successfully flash a 6.10.3 backup after this occurred with 6.11.0. Any suggestions on where to start troubleshooting would be appreciated. I didn't notice any relevant logs but I have a full copy of the flash drive archived. edit1: Testing with a clean 6.11.1 USB (all default settings) results in the same error, so it seems to be an issue with 6.11.1 and my hardware rather than my existing configuration. edit2: system specs: AMD A8-7600 Radeon R7 ASRock FM2A88X-ITX+

Having a similar as described above. Diagnostics attached. Hit start, Unraid progress animation shows for a few seconds, page refreshes. Status never changes from 'Stopped'. Ah, figured it out. Plugin will not start if "Enable for Cache and Pool Disks?" is set to "Yes". Seems to be working if toggled off. box-diagnostics-20220317-1251.zip