[solved] PHP errors on dashboard as log fills up

[34ty9unvf2]

I've had this issue in worse states but it seems to have returned today as I have left my unRAID install alone and worked within my VMs. I've attached the screenshots of the errors I see and the diagnostics file. When I have these errors I cannot access my Docker or VM controls within the Dashboard. Interestingly, the Dashboard also says my flash drive has been blacklisted. After a restart it seems happy again. I was able to shut down the specific docker containers from within their own respective UIs, and the same for my VMs. So the elements outside the Dashboard are working properly. However, in the past when the log fills to too large things will start crashing in the Docker containers and the unRAID UI will be a mess of php errors. I am running 6.7.0 as the 6.7.1 update basically killed my server (frequent complete crashes) until I rolled it back. Besides being unable to get video out from any GPU (tried recent nVidia and AMD) on any VM (tried Windows 10 and a Ubuntu 18.07?) things are working well. I recently added a few HomeAssistant docker containers though doing that within unRAID has been abandoned in favor of using a Raspberry Pi 3+.

So... lots of things going on here. I see the log files are full full full. Anything I need to change here so I don't have to reboot every month?

unraid-diagnostics-20190721-1648.zip

[trurl]

Put your flash drive in your PC and let it checkdisk. That is the most obvious problem you have at the moment. Be sure to use a USB2 port to boot from.

 

 

[34ty9unvf2]

I did chkdsk from the Windows 10 command line (pointing to the appropriate drive):

C:\WINDOWS\system32>chkdsk d:
The type of the file system is FAT32.
Volume UNRAID created 5/28/2019 4:00 PM
Volume Serial Number is 9603-4408
Windows is verifying files and folders...
File and folder verification is complete.

Windows has scanned the file system and found no problems.
No further action is required.
   16,367,616 KB total disk space.
          200 KB in 3 hidden files.
          208 KB in 26 folders.
      191,696 KB in 121 files.
   16,175,504 KB are available.

        8,192 bytes in each allocation unit.
    2,045,952 total allocation units on disk.
    2,021,938 allocation units available on disk.

Found nothing. The GUI version said there are errors when I started it, so check to fix errors. Checked... it found no errors. Any other ideas or a different check? I am using a USB 2 port on the server but its the front panel one. There are no rear USB 2 ports, only USB 3.

[34ty9unvf2]

I logged in to unRAID to see if its okay and I have the same errors, same non access to Docker and VMs. System is stable though besides unRAID's Dashboard having a fit. This "notifications disabled" is a reoccurring problem that coincides with other visible errors

[trurl]

Have you setup Notifications (you should)? If not, then you are going to get that message about Notifications.

 

Your diagnostics definitely had a problem with flash. It wasn't even accessible when the diagnostics were taken. Possibly there is nothing wrong with flash or the files on it, but rather some hardware problem causing flash to disconnect. Try another port. Is there only the one USB2 port? Usually they come in pairs.

[34ty9unvf2]

Yes, the notifications are normally set up. But the error about them not being set up seems to come with the inability to access my Docker or VMs. I don't even have anything auto-run (and nothing manually started) for Docker right now. Anyway, I do have another USB port that was up until last week populated by my Z-Wave USB transceiver that was just moved to the Raspberry Pi. So I moved it there and restarted. Will report. In the mean time I took another diagnostics file with a much smaller log, in case it might be helpful.

unraid-diagnostics-20190721-2138.zip

[trurl]

That diagnostic is no better. Flash disconnected. What kind of flash is it?

[34ty9unvf2]

That would probably be because I moved it to the other port prior to the diag and reboot. It's not complaining right now after 47 minutes uptime but I'll mess with it.

Edited July 21, 2019 by 34ty9unvf2
clarify

[JonathanM]

3 hours ago, 34ty9unvf2 said:

That would probably be because I moved it to the other port prior to the diag and reboot.

Yeah, don't do that. The unraid flash must stay connected continuously while the server is powered on. Otherwise you will see some strange behaviour.

[34ty9unvf2]

After a day and a half I'm not getting any visible errors. It seems the USB 2 port was at issue. I should probably look into getting an internal header to USB-A female cable in case the other one goes bad.