Would also much rather use Xen since I have that working and I'm not tied to other people,  I can just use standard distributions of the software I use.

Simple for me. Docker is plugins on steroids. VMs add new capabilities and extends unraid beyond Soho Nas appliance.

 

At the moment all I need is a Nas with plugins. Docker is all I need for now.

Would also much rather use Xen since I have that working and I'm not tied to other people,  I can just use standard distributions of the software I use.

 

+1

 

if Xen goes away in the 6.0 release, I'll stay on Beta 5.  I do not want to move my Debian Xen VM to KVM or build a new KVM from scratch.  I do not want to move the VM apps (OwnCloud, LMS, AirPlay, my own website, etc) to Dockers.  The Debian VM I am running on 6b5 is working perfectly and I can keep it updated very easily for security fixes with "apt-get upgrade".  OwnCloud and my website in the Debian VM are exposed to the Internet so security updates are important.  The process of updating multiple Dockers to keep security fixes current is not my idea of a good use of my time.

 

When 6b5 was released, I wasn't too excited about VMs and didn't really see the value.  I loaded IronicBadger's OwnCloud Debian VM just to see what it could do and found out that OwnCloud could be a real benefit to my business.  We now use it for centralized contacts, calendars, and collection of pictures from field smart phones.  I own a Handyman business and the Craftsmen in the field can see their scheduled work, look at the job pictures taken by the Estimator before they go to the job, and can take pictures on the job that are sent automatically to the OwnCloud server.  This technology has taken us to a new level of productivity.  Because of the ease of updating the Debian VM, maintenance for me is a breeze.

Yeah I'd prefer to stick with Xen myself too, but I'm not willing to forgo other unraid updates just for the sake of Xen.  If forced to I'll move to KVM if that is the VM environment that will get the most support from Limetech. Hopefully the Xen problems of beta 6 get fixed in beta 7.  If not, I'll start anew.

FWIW, I've messed with XEN and KVM now and while XEN's config file is much simpler, a KVM.xml is still do-able. If unRAID gets a built in VM manager like "Virt-Manager" it will be super easy for just about anyone to build a VM.

my experience with xen has not been 100% happy to be honest, i (and others) had issues with complete crash and halt whilst running b4 and b5 (reported to jonp and tom) , i then made the bold move of moving to b6 and switched to kvm (windows vm only) and have to say its been rock solid since, i have no idea if xen is more stable in b6 but kvm def is working very well for me.

I was using Xen for two VM's before b6 came out (one for MediaBrowser under fedora, another for Plex under fedora), but switched to docker.  My unraid box isn't directly exposed to the 'net, so that aspect of security is a non-issue for me.  Now that I've switched to docker, I'm running:  couchpotato, deluge, headphones, lazylibrarian, maraschino (for plex), mediabrowser, nzbdrone, nzbget, PlexMediaServer, and two others (sickbeard and sickrage), but they aren't always running as I've switched over to nzbdrone.  The one thing I've not got running under a docker is the mediabrainz server.  I'm currently running it in a vmware vm on my windows box.  If I can muddle through the complicated setup of it on the docker version, I'll migrate it to the unraid box as well.  There's way less of a space requirement for a docker of an app than a full vm for each (could always run most of that on a single vm install, but it defeats my purpose of having each app in it's own separate space/distro).

 

The only thing I'm running into is some of my docker'ed apps use quite a bit of CPU time, so it's looking like a CPU upgrade is in my near future...

 

I've got a few demanding apps running under Docker and windows VM's under KVM - everything is rock solid and haven't had to reboot unless I wanted to reboot, not true with Xen.

Hi Guys!

 

I have a problem/bug with this Unraid version.

 

Firstly I thought that it has something to do with KVM and all the VM stuff.

But yesterday i had the error without any VM running, just plain unraid!

 

So what is this error?

I don't know. But suddenly Unraid does not respond anymore. So the webgui, ssh, ntfs etc. are all dead.

When i press the powerdown button, nothing happens.

 

the only thing is, to do a hard reset

 

Then unraid will boot again, and everything seems back to normal.

I'm a computer guy, so i have a kind of knowledge about them. But here i am lost.

Also these Info are quite rare...

 

All i was doing, was playing a movie (rapsbmc) and after 70min the movie stopped und unraid was lost

 

I will try to recreate the error and trace a syslog.

 

My system config is:

 

CPU: AMD 5350

Mobo: Asus-AM1M-A

RAM: 4GB (2x2GB)

SATA: Digitus-DS-30104

DISK:2x 3TB WD Green, 2TB Samsung:

CACHE: 200GB FUJITSU

 

Unraid pro 6b6

 

any more Infos needed?

 

Regards

Owel

 

Edit: I did already memtest and prime, no errors so far ...

 

I have had the same issue several times.  I have to unraid servers running, and the one with dockers has never crashed, but my smaller one does occasionally when I copy a large amount of data to the array.  I could never really find anything because when it hangs, the screen is black, and upon reboot I don't see any errors on anything...  I haven't lost anything other than having to spend 10 hours or so with a parity check.  I was hopping the next beta might fix the issue since I can't really pinpoint what is causing the issue.

I am seeing some weirdness with highwater in the array.

 

In past versions it would fill all the disks to 97% before I added a new one. This time it is filling them to 99%.

 

/dev/md1        2.8T  2.7T  53G  99% /mnt/disk1

/dev/md2        2.8T  2.7T  52G  99% /mnt/disk2

/dev/md3        2.8T  2.7T  94G  97% /mnt/disk3

/dev/md4        2.8T  2.7T  114G  96% /mnt/disk4

/dev/md5        2.8T  2.7T  120G  96% /mnt/disk5

mine is even 100%...

 

root@Tower:~# df
Filesystem      1K-blocks       Used  Available Use% Mounted on
/dev/sda1          503016      95600     407416  20% /boot
/dev/md1       2930177100 2900943776   29233324 100% /mnt/disk1
/dev/md2       1953454928  638203768 1315251160  33% /mnt/disk2
/dev/sdc1       195360952  113716332   80569332  59% /mnt/cache
shfs           4883632028 3539147544 1344484484  73% /mnt/user0
shfs           5078992980 3652863876 1425053816  72% /mnt/user

 

can this be causing my error?

Mine is pretty accurate..

 

/dev/md1        3906899292 2471883712  1435015580  64% /mnt/disk1

/dev/md2        1953454928  380792264  1572662664  20% /mnt/disk2

/dev/md3        1953454928  379507956  1573946972  20% /mnt/disk3

/dev/md4        1953454928  382863832  1570591096  20% /mnt/disk4

/dev/md5        3906899292 2334560228  1572339064  60% /mnt/disk5

/dev/md6        1953454928  381260860  1572194068  20% /mnt/disk6

/dev/md7        1953454928  378405016  1575049912  20% /mnt/disk7

 

Do you have a share that has the split level  missing?

I've been an Unraid user for a little over 2 years.  I have fallen in love with V6. Not so much Beta 6.(network bug bit me)  I am a Linux System Admin by trade and I am much more comfortable building VMs then using docker containers.    One of the main reasons is that I know exactly what packages I want to install for my linux VMs and I am only installing what is needed to make my application run.  All of my python based apps run in and Ubuntu LTS VM and just about everything else runs in a CentOS VM.  I love the flexibility that the new Xen kernel provides.

Several of us have noted that their unRAID beta hangs when completing the powerdown -r command (part of APCUPSD);  the hang comes early in a reboot leaving a "SYSLINUX...." text on the screen and no cursor.  Usually, but not always, the server will boot normally if the power button is pressed (more on that later).

 

I'm using a new ASRock AD2550R motherboard and a flash drive purchased from LimeTechnology with unRAID 5.05 on it that was subsequently upgraded to 6b5 and now 6b6, I boot without xen in 6b6.

 

Experiment

 

In addition to failing on the powerdown script I tried the stock 'reboot' command (after unmounting the array) and this hangs too.  I used the 'reboot' command for most of the tests below, since the APCUPSD powerdown script was not available.

 

Many of the threads with this kind of issue suggest that the flash might be corrupted.  I tried copying all new OS files (except config) and running make_bootable_mac.  Reboot still hung at the same place.

 

I shut down the server and disconnected all of the drives.  My boot flash failed again.  I tried it in all three USB buses on my MB.  I tried every setting in BIOS I could find. Still hung.

 

Then I took a brand new Sony flash and ran these trials:

 

unRAID 5.05 downloaded today: boots and 'reboot's fine.

 

unRAID 6b5 downloaded today: boots and 'reboot's fine.

 

unRAID 6b6 downloaded today: boots but hangs on 'reboot'.  Again this is on the stock unregistered unRAID, freshly loaded and freshly "make_bootable"ed on a new flash drive.

 

I tried using the stock powerdown script in that stock install but I couldn't figure out the switch than might cause it to reboot instead of just power down.

 

TL/DR:  Reboot is broken somewhere between 6b5 and 6b6.

 

Now some other hinkyness:

 

The reboot hang seems to happen if the re-power happens immediately upon depowering.  If I shut down the server with a three-second powerbutton push and then immediately push the powerbutton to restart, it (usually) hangs.  If I wait as little as 5 seconds to push power, it (usually) boots without hang.

 

This has made me think that a workaround might be possible if the motherboard could delay a few seconds before repowering, but I can't see how to do that, though there are some references to a USB Mass Storage Boot Delay on the web, usually because it makes your reboot too slow.

 

TL/DR:  The problem might be corrected by delayed repowering after shutdown.

 

I'm not sure where to find a log of the early boot process; it seems unlikely to be recorded.

 

While most of us have lived with this, it has consequences for unattended servers or for recovery after power failure.  I hope that it can be tracked down and killed.

 

 

Dennis

 

What version of Powerdown are you using?  Look in your log after a reboot and see what version of Powerdown is being installed.  The older Powerdown has a problem with 6.0.

Just wanted to confirm something for those that are wondering, Xen is NOT going away in beta 7.  We have it in there.  We have tested with it as well and everything works ok for us across 3 different test systems we have.  That said, the primary bug in beta 6 I believe with Xen has to do with the xenet skb rides the rocket issue.  Network in general has been "unstable" with Xen and there was nothing we could do to fix it.  In beta 7's internal testing thus far, we have not run into any Xen networking issues yet, but again, we didn't have them on our test systems with beta 6 either.

 

There was reference to some patches for Xen in the 3.16 kernel which is what we are using for beta 7.  Once released, we are looking for feedback from folks to let us know if the issues persist or if they've improved.

 

Thanks everyone...back to work on beta 7...

What version of Powerdown are you using?  Look in your log after a reboot and see what version of Powerdown is being installed.  The older Powerdown has a problem with 6.0.

 

I updated Powerdown with the beta.

 

Aug 17 11:20:08 Tower logger: Executing install script for apcupsd-3.14.10-x86_64-1_rlw.txz.
Aug 17 11:20:08 Tower logger: Package apcupsd-3.14.10-x86_64-1_rlw.txz installed.
Aug 17 11:20:08 Tower logger: 
=
Aug 17 11:20:08 Tower logger: plugin: skipping: /boot/packages/powerdown-2.06-noarch-unRAID.tgz already exists
Aug 17 11:20:08 Tower logger: plugin: running: /boot/packages/powerdown-2.06-noarch-unRAID.tgz

 

Seems to be the newest one on github.

(paranthetically, is this calling for a reinstall of apcupsd on every boot?)

 

Remember too that the issue seems to be with reboot (too?).  Not sure how Powerdown calls a shutdown/reboot.

 

Thanks

D

...(paranthetically, is this calling for a reinstall of apcupsd on every boot?)...

The unRAID OS is unpacked into RAMfs from bzroot. Everything that is not in that archive is reinstalled on every boot.

Just wanted to confirm something for those that are wondering, Xen is NOT going away in beta 7.  We have it in there.  We have tested with it as well and everything works ok for us across 3 different test systems we have.  That said, the primary bug in beta 6 I believe with Xen has to do with the xenet skb rides the rocket issue.  Network in general has been "unstable" with Xen and there was nothing we could do to fix it.  In beta 7's internal testing thus far, we have not run into any Xen networking issues yet, but again, we didn't have them on our test systems with beta 6 either.

 

There was reference to some patches for Xen in the 3.16 kernel which is what we are using for beta 7.  Once released, we are looking for feedback from folks to let us know if the issues persist or if they've improved.

 

Thanks everyone...back to work on beta 7...

 

Thank you for the update.  I had the network issue pretty quickly in b6 when I had heavy network traffic to my Owncloud Debian VM.  I'll give it a go once it is released

Thank you for the update.  I had the network issue pretty quickly in b6 when I had heavy network traffic to my Owncloud Debian VM.  I'll give it a go once it is released

 

Same here - the network on my Xen ArchVM fails within five minutes of startup when running b6.  I'm looking forward to testing with b7!

So this network issue only affects those that run VMs on their unRAID host?  If I run unRAID simply as unRAID and don't use Xen then networking is fine?

So this network issue only affects those that run VMs on their unRAID host?  If I run unRAID simply as unRAID and don't use Xen then networking is fine?

 

That is what I understand.  I believe that those who have moved their VMs to KVM do not experience problems.

So this network issue only affects those that run VMs on their unRAID host?  If I run unRAID simply as unRAID and don't use Xen then networking is fine?

 

That is what I understand.  I believe that those who have moved their VMs to KVM do not experience problems.

Excellent.  I use Xenserver as my actual VM hosts but I'm keen to try more of v6 but didn't want it to compromise the network to unRAID.

I see the same issue.  I just hopped between 6.06 and the latest 5. 6.06 hangs after a reboot.  I need to power cycle.