Ive not fallen fowl of this until now, just upgraded from rc3 to rc4, rebooted the system and it looks like sabnzbd and jenkins docker containers are having connectivity issues to the internet, interestingly both can do name resolution, i can access sabnzbd docker container webui over the internet, as well as locally.

 

jenkins is showing the following message in configuration of a job "java.net.UnknownHostException: github.com" i can ping github.com from the running container, and i can use curl to view the page source for github.com but it seems to think there is an issue, this COULD be a specific issue with a jenkins plugin im using, but i have had this running previously and this is the first time ive seen this message.

 

sabnzbd is just basically refusing to download anything, nothing in the logs, i have restarted the container but nothing, i am aware of the trick of switching from bridged to host and back but wanted to see if there is anything else you want me to try before i do this.

 

I am running a windows 8.1 vm, and this is set to start on boot, its also using the older "xenbr0" bridge as i need it to pickup a dhcp lease from my router.

 

Currently i have left the dockers in this state to try and help Tom/Jonp do some further diagnostics with me if they want, please contact me asap if this is the case, i can of course send you the diagnostics tools generated zip (awesome addition guys!!) and any other files you might require, this is my live system so i obviously don't want it left in this state for too long but im keen to help you guys out if i can.

 

edit - damn looks like i accidental fixed sabnzbd, i figured i could shutdown sabnzbd and create a new container thus giving me a working system whilst keeping the old screwed one for further diags, unfortunately my plan didnt work as i forgot to change the container name so it updated the existing container (with the same configuration) which interestingly then resulted in it working!

Right this is an extremely peculiar bug.  It's a fairly complex networking setup now.  If anyone can make it happen "on demand" then please post here.  Meaning, "If I do this, then it always happens without fail that network connectivity is lost" - what is this?

damn looks like i accidental fixed sabnzbd, i figured i could shutdown sabnzbd and create a new container thus giving me a working system whilst keeping the old screwed one for further diags, unfortunately my plan didnt work as i forgot to change the container name so it updated the existing container (with the same configuration) which interestingly then resulted in it working!

 

That was my experience, never did any of the changing host to bridge and back again or vice versa, just edited the config and saved without changing anything and that fixed the issue for me and haven't seen it since, despite many reboots.

Since this might take a little while to solve, here's the a method to issue commands from the container's "command line." It uses the "docker exec" command. I used it to test a container's DNS resolution by using the ping command. I'm sure there are other uses in troubleshooting Docker containers.

 

http://lime-technology.com/forum/index.php?topic=36696.msg340907#msg340907

Since this might take a little while to solve, here's the a method to issue commands from the containers "command line." It uses the "docker exec" command. I used it to test a container's DNS resolution by using the ping command. I'm sure there are other uses in troubleshooting Docker containers.

 

http://lime-technology.com/forum/index.php?topic=36696.msg340907#msg340907

 

A tip is to add the environment variable TERM with the value xterm in the docker container advanced settings.

That way you can run 'top', 'nano' and more, without getting the "TERM environment variable not set." error message.

Right this is an extremely peculiar bug.  It's a fairly complex networking setup now.  If anyone can make it happen "on demand" then please post here.  Meaning, "If I do this, then it always happens without fail that network connectivity is lost" - what is this?

 

Hi Tom, I can't offer you this unfortunately but I can send you a copy of my syslog if it's of any use?

This is most likely just dns resolving issue now (the race condition one), right?

I think after i updated my unraid to b14 or rc2, and set my sabnzbd to numeric server ip, i've not seen any network related errors.

This is most likely just dns resolving issue now (the race condition one), right?

I think after i updated my unraid to b14 or rc2, and set my sabnzbd to numeric server ip, i've not seen any network related errors.

 

If you use static DNS IP addresses ("Obtain DNS Server Address Automatically: No" and specify IP of your DNS server) I don't think you will be affected by this issue.

This is most likely just dns resolving issue now (the race condition one), right?

I think after i updated my unraid to b14 or rc2, and set my sabnzbd to numeric server ip, i've not seen any network related errors.

 

If you use static DNS IP addresses ("Obtain DNS Server Address Automatically: No" and specify IP of your DNS server) I don't think you will be affected by this issue.

 

Folks, I'm marking this one solved.  If anyone else has an issue with this, please report back here, but both feedback and testing from us and the community has shown that rc6a has this issue fixed.  Thanks for your patience and assistance testing with us to help bring this issue to a full close.