it seems to shutdown unRAID, since I can't telnet, or see any drives or shares, but the server continues to run until I actually hit the off button on the box.

 

This seems like an error to me.

Do you mean 'shutdown'?  Before you can select shutdown you have to 'stop' the array - it is at this stage that the drives and shares will become inaccessible.

 

After you hit the off button does the array do a clean start, or does it perform an automatic parity check?

Yes, after stopping the array, you can still see the flash drive, and also can telnet into unRAID, but once I do 'shutdown', I can't do either of those, but the machine never actually turns off.

 

When I started it earlier tonight, it did not start a parity check, so it seems to have been 'clean', but I have a separate red-ball issue, which I think prevents a parity check anyway, so I'm not positive it was a clean shutdown.

Maybe emhttp got OOMed? Not sure this is likely on 64bit unRAID though. One of my first experiences with unRAID was the web interface "crashing" while trying to shutdown.

I don't know; how could I test this theory?

You can see if emhttp is running with

ps aux | grep emhttp

 

You should get a line like

root      2470  0.0  0.0  88908  3336 ?        Sl  Aug15  0:25 /usr/local/sbin/emhttp

 

but I can't get into telnet after hitting 'shutdown'.  it acts as if it's shutdown, but the machine just keeps running.  unraid is not running, but the machine does.  that's the problem.

How about the console? Do you have a monitor and keyboard attached? Maybe you could copy the syslog to your flash drive and post it.

no, neither monitor nor keyboard attached currently.

 

I'd messed about with getting GPU passthru to windows a while back, and since then, I can't see the console after about halfway thru bootup.

 

I'm sure I can get that squared away, but I'm fighting with a redball situation right now, and I'm going to get that fixed up before making changes to the settings I have for trying passthru, since beta 7 might make it all work, so I want to try that before changing things.

 

I'm hoping I my have time this weekend to more dedicated effort on unRAID.

Maybe your emhttp has gone unresponsive.

You can try the clean powerdown script. It uses another method to cleanly powerdown your server.

Maybe, but even so, the bug should be fixed, no?  If the built-in shutdown procedure doesn't work, that's just not a good thing.

 

Once I get my other issues resolved, I'll look into the powerdown script, so I can just press the power button on the server, but that's probably gonna be a little while before I have time for that.

I think there must be something specific t your system?  I certainly find that on my system when I select shutdown on the GUI it results in the system being powered off.

 

I think having a console may be essential to try and help debug this system.    It sounds as if the closedown at the Linux level is not completing for some reason.

 

Do you have any plugins/docker containers/VM's running?    If so it might be worth seeing if you can reproduce the symptom with a completely clean system.

I'm sure it is specific to my system.  I have 5 dockers installed, but usually only 2-3 are running.  The only plugin I have running is the docker plugin to install/manage the dockers.

 

Once I get my other issues resolved, I'll dig deeper into this issue and see if I can isolate it.

 

I'm going to upgrade to beta7 for various reasons, and if I still have issues after that, I'll see what I can figure out.

I'm having the same issue with beta7.  Webgui shutdown worked fine in beta6. 

Anyone still having these issues in beta10/a?

I looked into this more closely the last couple days, and in my case, what happens is that the server shuts down, but then turns itself back on after a couple of seconds.  I checked my BIOS and I have it set to stay turned off after a power failure, but even if I didn't, i wouldn't think an unRAID clean shutdown should be recognized as a power failure by the computer.

 

in my testing, it seems that about 10% of the time, it shuts down and stays shutdown.  I'm at a loss as to why it turns itself back on most of the time.

 

But, now that I know, I just walk to the box, and turn off the power supply after it shuts down, preventing it from turning itself back on.

 

Weird. 

Sounds as if you may have the "Wake on LAN" option activated in the BIOS and therefore network traffic us turning the server on.

I do have wake on LAN enabled, but I didn't expect it to wake from an off situation, only if it were sleeping.

 

if i disable it, will clients be able to wake it from sleep?

I do have wake on LAN enabled, but I didn't expect it to wake from an off situation, only if it were sleeping.

 

if i disable it, will clients be able to wake it from sleep?

Soft off is pretty much just the deepest form of sleep. See here. WOL is still very much able to wake it, if it has been configured correctly.

Ah, interesting.  I think I have options for WOL from my BIOS, I'll have to see if I can enable it for everything but G5 (soft off), since that's what the GUI does, and when that happens, I want it to stay off, but want it to WOL from any other sleep state.

 

thanks for the info.

Ah, interesting.  I think I have options for WOL from my BIOS, I'll have to see if I can enable it for everything but G5 (soft off), since that's what the GUI does, and when that happens, I want it to stay off, but want it to WOL from any other sleep state.

 

thanks for the info.

 

Justin.  Let me know what happens from this.  I actually have seen this myself, but must have made a change that "fixed it" from happening again, but don't know what it was.  Would like to know for certain if this was just a rare oddity or part of a bigger issue.  Thanks!

... what happens is that the server shuts down, but then turns itself back on after a couple of seconds.

....

in my testing, it seems that about 10% of the time, it shuts down and stays shutdown.

 

Mine does this, but very occasionally - like, maybe, 5% of the time.  So, 95%, good, 5% bad ... unlike you with 10% good 90% bad.

 

But, now that I know, I just walk to the box, and turn off the power supply after it shuts down, preventing it from turning itself back on.

 

It's not just webgui shutdowns - I've seen it happen on an aocupsd shutdown too, but I don't worry too much because the ups turns off before the disks remount.

 

Wierd indeed!

Can you guys post copies of your syslinux.cfg file contents up here?  Also any pictures from power settings in your BIOS may also be helpful to examine here.  Thanks!

I"ll try to get a shot of m BIOS soon, but my syslinux.cfg is below...

 

default /syslinux/menu.c32
menu title Lime Technology
prompt 0
timeout 50
label unRAID OS
  kernel /bzimage
  append initrd=/bzroot
label unRAID OS Safe Mode (no plugins)
  kernel /bzimage
  append initrd=/bzroot unraidsafemode
label Memtest86+
  kernel /memtest
label Xen/unRAID OS
  menu default
  kernel /syslinux/mboot.c32
  append /xen --- /bzimage --- /bzroot
label Xen/unRAID OS Safe Mode (no plugins)
  kernel /syslinux/mboot.c32
  append /xen --- /bzimage --- /bzroot unraidsafemode

Did anyone ever find a fix for this issue - the computer powering itself back on after a shutdown?

 

I still experience this on occasions

 

I have looked at BIOS setting (SuperMicro X9SCM-iiF) and I don't believe that I have any 'wake on' settings enabled.