Unraid unresponsive ... not sure why?

[montery]

My unraid became unresponsive, and by looking at the logs, it looks like it ran out of memory?  In order to get the webgui back, I had to reboot.  I was not able to shutdown gracefully.  I have upgraded from V5 to V6, and it died when running V6.01.

 

I have 16 gigs, was running a Windows 7 VM, and numerous plugins.

 

I've attached the diagnostics file, which includes the last log file prior to re-booting.

 

Any recommendations?  Should I get more ram??

 

Thanks!

tower-diagnostics-20150807-2216.zip

[BRiT]

I see youre running Dynamix System Stats plugin. Make absolutely certain you increase the size of /var/log to at least 384 megs or else you will absolutely run out of space there. There is a command you need to put in your go file to do this.

 

Second, at some point you must have had system crash, as there is a warning about the filesystem on your usb thumb drive about possible filesystem corruption, so you should run a check filesystem on it. I think people typically do that by running it under Windows. This is your san disc cruzer device.

[BRiT]

You also had several issues of usb2 device resets, you may want to try using a different usb port.

 

 

You also had instances of your disks /dev/sdb and /dev/sdg failed to read anything at all when you started a parity check. It could be cables or it could be controller related. This started on August 3rd.

 

Im still scrolling through the log file, but theres a MFT of read errors to get past before i can see if anything else happened.

[BRiT]

Later on it looks like your cache drive might have filesystem corruption in your /Downloads directory, as the mover died at that point on August 7th throwing a system stack trace. After this stack dump the system hit out of memory conditions and started killing processes. I have seen other RFS corruptions leading to out of memory conditions. So do a check filesystem on your cache drive too. This will be from unraid command prompt or from the gui if there's a way of doing that in dynamix.

[montery]

BRiT,

 

Thank you.

 

I did some googling against the Unraid Forums, and yes, it looks like var/log needed a big boost.  I upgraded it to 1024mb, rather than your suggested 384mb.

 

For benefit of others, the command I added to my go file (and ran on the console) was:

 

mount -o remount,size=1024m /var/log

 

 

Additionally, /dev/sdb and /dev/sdg are some ancient drives, so I've removed them from the array.

The Sandisk Cruzer USB flash drive is an antique -- it is the same flash drive I used when I first bought Unraid hmm... 10 years ago now??   I'll buy a replacement and then bug Tom about the new license.

 

Lastly, the /Downloads directory is a user share, not on my Cache drive.  However, I may have had a Downloads directory in the cache drive previously.  Anyhow, according to this Wiki entry, it looks like BTFRS can't be fixed, just yet, so I'll wait and see.

 

Otherwise, I'm hopeful that updating the /var/log size will prevent the GUI from freezing.

 

Next up, finding out why my shutdown@ script on the flash drive didn't work!

 

Thanks again, BRit, I really do appreciate your help.

 

 

[itimpi]

Bear in mind that increasing the size to 1024MB means that you have dedicated 1GB of RAM to holding log files.  This may well be more than most people are prepared to dedicate to this use. 

 

A better solution is to limit the amount that gets logged in the first place as logging that is too verbose is normally not of much use as it is often too much effort to look through very large logs.