[SOLVED] v6.8.2 - Strange WebGui behavior then becomes unusable, part 2

[The_N4RF]

Hi Everyone, Still dealing with a very frustrating issue.  My unraid server (onprem) continues to become unresponsive after about 24 hrs of uptime.  Both the webviewer and the CLI seem to enter a semi-crashed state.  The server stops responding to shutdown commands (via web, console, terminal), so I have to hard-boot the server to get it to work again.  Some of the initial symptoms include nonsense cpu load reports in the web viewer, and the "unnassigned devices" in the Main section never loads.

 

Eventually the web viewer stops loading completely and I'm just greeted with this:

 

Has anyone seen this?  This is new behavior for me since 6.8.

onprem-diagnostics-20200224-1309.zip

[somebuddy]

In my case such a behavior was caused by a faulty RAM. Please try a MemTest to check your RAM.

[The_N4RF]

Thanks for the idea.  I may still be stuck even there.  I plugged a monitor into my tower and started it up.  At the boot options I chose "memtest86+".  The command line responded "/loading memtest... ok", then the machine just rebooted and brought me to the same boot options screen.  A few attempts yielded the same results.  Another hint is greatly appreciated.  I'm still investigating.

[The_N4RF]

Mmhmm.  Got it working by unplugging an external usb3 hdd before booting.  I'm using all of my sata ports right now, so I had one unnassigned device plugged in via USB just for writing my security footage to.  Didn't want to ruin my parity disk by having to keep up with a 24/7 stream of video data.  Maybe that's a bad idea.

[testdasi]

2 hours ago, The_N4RF said:

Mmhmm.  Got it working by unplugging an external usb3 hdd before booting.  I'm using all of my sata ports right now, so I had one unnassigned device plugged in via USB just for writing my security footage to.  Didn't want to ruin my parity disk by having to keep up with a 24/7 stream of video data.  Maybe that's a bad idea.

USB storage devices shouldn't be used on an on-going basis.

Your unresponsiveness could be IOWait as the CPU waits for the USB device to respond and if the device is not working well, it would lead to very long wait. Given you are writing to it 24/7, the USB controller could very well have overheated and is on its way out (if not already effectively dead).

[The_N4RF]

Seems like a very likely scenario.  The RAM test is proving successful, so I'm going to try to reconfigure the server to use a sata drive for the video streams and see if my uptime improves.

 

Thanks for the help!

[testdasi]

On a side note, if you run out of SATA ports, it is generally better to replace low capacity drive with higher capacity drive rather than adding more drives.

HDD fails in statistical patterns so the more drives you have, the more likely for you to have a failed drive.

[The_N4RF]

Yes, but I can get those low capacity drives really cheaply :).  

 

Removing the USB drive from long-term use appears to be the solution.  I'm using a SATA drive as an unassigned device now and the server appears to be back to 100%.  Thank you!

[The_N4RF]

Update: I was still dealing with some crashing.  Realized my trusty boot usb stick was failing.  Deployed to a new stick and now my server has been running 48 days on the trot.