Hello all,

 

For the purpose of this conversation i'll post an outline of my setup before explain my issue

I have unraid running xen with three pv guests running ubuntu server 12.04.4 LTS. These three hosts live on a drive outside of my array formatted with ext4 which is mounted to the unraid machine via the go script.  The three guests which i have setup are one host for sab one host for sickbeard and one host for plex.

 

To achieve the integration between applications require, I have created two NFS shares on my unraid box (one cache share-cache drive) and (one media share-this is on the unraid array).

 

All three virtual machines mount the relevant NFS share on startup to the unraid box via fstab on boot to provide their relevant network shares they use (eg sab gets access to the cache nfs share where as the sickbeard hosts gets both mounted and the plex only gets the media share mounted). When I boot my unraid box all VM's start-up (using autostart in the webgui) and run perfectly (eg access all the web interfaces work and work as required).

 

The issue is when as soon as data is transferred between the vm host and unraid box, the particular host competing the data transfer loses network connectivity and requires a reboot for it to be brought back on-line. This problem isn't limited to one vm host it will happen whichever host begins to perform significant data exchange (more than a few 100mb roughly)

 

Is anyone experiencing issues with Unraid beta 6 as well ? Where their vm hosts lose network connectivity ?

 

BTW i am somewhat of a novice so please excuse me if havn't explained anything properly.  Thanks in advance for any feedback or suggestions.

 

I was experiencing something similar where my ArchVM in Xen would lose connectivity at random. Sometimes it came back, other times I had to reboot. I didn't find a solution and ended up moving my apps to Docker which eliminated the issue.

 

As a side note, there is no reason to run 3 VMs for 3 apps. You can run all of those in a single VM, along with others (I had sab, sb, cp, maraschino, mariadb and plex in my VM). You've severely over complicated your environment beyond anything necessary.

 

If you have a cache drive, or another external drive you can mount with BTFRS I would highly recommend going to Docker. It's far easier and less overhead.

Thanks for the feedback, I take your point re complexities  of my setup. The reason for the three separate vm's was more for me just to play around. I guess i need to look further into docker.  I do believe the issue i am experiencing and from the sounds of it you did as well might be an issue with xen implementation or something which might be worth while being explored.

 

Thanks again

Benny

Is anyone experiencing issues with Unraid beta 6 as well ? Where their vm hosts lose network connectivity ?

 

Yes!  I don't think that I', the only person to revert to beta5a because of this issue with beta6.

Is anyone experiencing issues with Unraid beta 6 as well ? Where their vm hosts lose network connectivity ?

 

Yes!  I don't think that I', the only person to revert to beta5a because of this issue with beta6.

 

Oh ok so you also experienced the issue I outlined above and were able to resolve it through downgrading to beta 5a ? I might give that a try that before moving to the docker solution.

 

Thanks again for the feedback

I can confirm by downgrading from unraid version 6 beta 6 to beta 5a and changing nothing else that my problem is now gone.

 

Thanks for those who provided feedback.

Did you ever look at the console to see if there were errors? There has been an issue rolling around that sounds like this..the "skb rides the rocket" problem.

 

I upgraded to V6 beta 7 and it was fixed. Didn't look into the errors much

I upgraded to V6 beta 7 and it was fixed. Didn't look into the errors much

 

Have you followed issues with RFS in beta 7 and 8? You need to upgrade to beta9 ASAP!