6.0rc3: unresponsive after requesting array stop. (FIRST TIME ONLY)

[johnny121b]

[glow=red,2,300]UPDATES at end of posts.  5/20/15.[/glow]

 

I'll preface this by saying- This is a test setup for me- not my real server.  I don't need help making this specific configuration work, but I (am) concerned that I was so easily able to derail Unraid with such a minimal level of activity.  And I'm wondering if this is indicative of an underlying instability or if there's just some fundamental use error that I don't understand- that unraid hasn't protected/prevented me from.

 

I decided to play around with 6.0rc2 this evening, and I'm disappointed to say it only took me 15 minutes to find one of 5.0's annoyances is still around- maybe even worse.  Namely- an endless series of Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)... when I attempt to take the array offline.

 

The difference is:  (1) I didn't yet have any user shares. (2) The endless series stopped, but the http interface was unresponsive (3) The server screen said Error: Device scan failed. 'mnt/disk1/docker.img' - Block device required.

 

I used the shutdown -r now command via telnet to restart the server, but I want to ask.....WHY is unraid unable to keep control of its own house even when using no plugins?  (Actually, I had ONE plugin installed at that point- gfjardim's 'community repositories')  And I had two dockers- Sabnzbd & Sickbeard....neither was configured & functional, just 'started'  My frustration is that I don't see why I lost control of unraid while almost nothing was running.  I could sorta understand if I were running Sabnzbd, Sickbeard and PLEX....but....Sabnzbd wasn't set up, nor was Sickbeard....AND they were both more-or-less idling in Dockers.  Freshly formatted flash drive, one data drive in the array, using a 30-day license just to experiment.... 

 

I thought the switch to Dockers was supposed to allow Unraid to better maintain control of things.  All I did was ask to take the array offline, and I feel like I just pulled the wrong block out of a Jenga tower.  (no pun intended)

 

What gives?  Did I do something so wrong that I should've lost control of the system...requiring a telnet rescue?   

[jonp]

I'll preface this by saying- This is a test setup for me- not my real server.  I don't need help making this specific configuration work, but I (am) concerned that I was so easily able to derail Unraid with such a minimal level of activity.  And I'm wondering if this is indicative of an underlying instability or if there's just some fundamental use error that I don't understand- that unraid hasn't protected/prevented me from.

 

I decided to play around with 6.0rc2 this evening, and I'm disappointed to say it only took me 15 minutes to find one of 5.0's annoyances is still around- maybe even worse.  Namely- an endless series of Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)... when I attempt to take the array offline.

 

Were you connected to the server via SSH or Telnet when this occurred?  Lingering sessions can have this affect if they are traversed to a path that is attempting to be unmounted.  Doesn't matter if you have user shares or not.  If you are logged in and under /mnt/cache, /mnt/disk#, or /mnt/user, you are going to get that message.

 

The server screen said Error: Device scan failed. 'mnt/disk1/docker.img' - Block device required.

 

This is interesting.  you didn't have a forward slash ( / ) before the mnt?

[RobJ]

Can you describe the steps you took to prepare the flash for v6?

 

It really sounds like there are vestiges of v5 on the flash drive, as there's no way it would be warning about user shares if nothing share related was there.

 

Apart from that, unRAID is not that fragile, and no one else is having the same problem, so must be a unique problem, hopefully simple and solvable!

[johnny121b]

jonp

Yes, I was logged in via telnet.  Strike 1.

Not sure about the leading "/"  I was typing from memory, spinning around to see the server screen from across the room- likely my error.

RobJ

I formatted the drive and unzipped 6 onto it, running the makebootable batch file...BUT I did a quick format....and even though I would normally thought that was enough,  AND now that you mention that, I DID recall seeing something that reminded me of a past unraid config I'd used that flash drive for.  So I think that's strike 2.

 

Thanks to you both for responding.  I'll try this again tomorrow- after I've had more sleep. 

[jonp]

jonp

Yes, I was logged in via telnet.  Strike 1.

 

If you were just logged in, that's one thing, but if you were inside a path to a disk/share that was attempting to be unmounted, then that is the cause of your issue.

 

Not sure about the leading "/"  I was typing from memory, spinning around to see the server screen from across the room- likely my error.

 

Chances are that didn't have anything to do with the attempt to stop the array.  The telnet session is the most likely culprit at this point.

[Wally]

johnny121b,

 

I have this exact same problem with RC1 and RC2. For some reason the cache drive disappears and cannot be accessed by Plex or KVM and when trying to restart to troubleshoot, I get the "Retry unmounting use share" errors since the shares located on the cache are gone. I tried shutdown -r now but nothing happens so I've had to hard reset which the forces a parity calc when restarted. I've gone back to the previous unRAID version for now.

[johnny121b]

Afraid I'm obviously no help, Wally.  I'm a minor menace within Linux.

 

I created some of my own problems, but that's exactly what I needed to do- to see just how robust and idiot-proof the new version will be.  I've used Unraid for years, and I never knew the telnet session would hang a shutdown.  Now that my server has grown significantly, I'm not willing to risk its data, so I always try out new software on a spare PC...and I never move to a new version until the dust settles and my backups are up-to-date.  That makes for less stress for me AND for those helping me.

 

I [do] hope my post helped underscore a probable expectation among Unraid's [non-Linux-guru] users, that Unraid always be in-control and responsive.....even if that means scolding the user for doing something incorrectly.  Don't give me a button to push- if I'm going to be punished for pushing it.  IF there's some reason OFFLINE cannot be chosen, prevent me from pushing it and guide me into completing the operation correctly. 

 

And I do think they're working towards that level of refinement.  I just hope they keep-in-mind.... their user may be their mother or father....who understand the WHAT, but not the HOW.

[johnny121b]

Update: 

 

Trying this again.  Much earlier in the evening.  This time with rc3.  This time with a flash drive that's been fully formatted fat32, with rc3 extracted onto it, made bootable.

 

 

Boot system.

1 4Tb data drive.

No cache drive.

Start array.

Turn on Docker.                    Set docker image to:  /mnt/disk1/docker.img  (Remember- This is just a test system.  I just wanna see things run & get a feel for it)

Add repository.                      https://github.com/gfjardim/docker-containers/tree/templates

Add Sabnzbd docker.            (Didn't log into it at all.  It's unconfigured and in whatever state the docker starts with)  Mapping:  /config = /mnt/disk1/sabnzbd    /downloads = /mnt/disk1/downloads

Go to Main

Choose "I want to do this"

Choose "Stop"

...Watch as the endless series of unmounting messages stack up.

telnet in and issue a shutdown now command

Watch as system closes things down and stops at tower login prompt.

Tap POWER button on PC.

Watch console as system does a FULL shutdown

Restart system and begin typing this post, feeling self-righteous

Try it again

And everything now works beautifully!!

I've spent the last 15 minutes continually restarting the system.  Goes online and offline easily.  Docker restarts.  Sabnzbd seems happy. 

But just so you know- that first time through...[glow=red,2,300]SOMEthing [/glow]was preventing the system from shutting down.

And I DO hope the 'docker repository' won't forever be something the user has to search for online.  Have some sort of default....to help new/inexperienced users get started. 

[ijuarez]

once your in this loop how do you stop it so you can shut it down.

 

i need add a drive?

[Squid]

And I DO hope the 'docker repository' won't forever be something the user has to search for online.  Have some sort of default....to help new/inexperienced users get started.

Complete List of docker repositories.  Continually updated: http://lime-technology.com/forum/index.php?topic=37958.0

 

Plug-in to really make docker's super easy to add: http://lime-technology.com/forum/index.php?topic=39106.0

[Gragorg]

I have had this exact same issue.  If you power down your server remove your USB flash and install it in your pc.  Mine automatically detected errors and fixed them.  Plug it back in server and reboot and Plex and everything returned to normal.  This is the third time in the last year this has happened to me.  I think my USB may be starting to fail.  Did you find a way to powerdown safely?