unRAID Server Release 6.0-rc6a-x86_64 Available

Download

 

Clicking 'Check for Updates' on the Plugins page is the preferred way to upgrade.

 

Finally I think this squashes the "docker dns" bug.

 

To restore proper operation of your containers:

 

1. Remove all places you might have added --dns=<ip-address> switch: docker.cfg file and/or "Extra parameters" on container edit page.

2. For each container bring up edit page, click save.  Note this will also "update" the image.

 

That's it, communication with outside world should be restored.

 

Bug was in a docker source file.

 

Changes
=======

Version 6.0-rc6a
----------------
- docker: fix bug in docker/daemon/container.go: /etc/resolv.conf permissions should be 0644 not 0600
- webGui: add 'restart' to docker context menu

 

quick check before applying this update..

 

 

if i previously removed the --dns options from my containers whilst testing the permissions thing, do i need to edit/save containers ?

quick check before applying this update..

 

 

if i previously removed the --dns options from my containers whilst testing the permissions thing, do i need to edit/save containers ?

 

Probably not but it doesn't hurt 

1st reboot and every one of my containers now has internet connectivity.

1st reboot and every one of my containers now has internet connectivity.

 

That's good, right?

1st reboot and every one of my containers now has internet connectivity.

 

That's good, right?

 

kind of , my dockers work but now i have nothing to bitch about, lol.

seriously though, damn good job team LT.

Looks good to me, but then I've never been bit by the bug.

Hopefully this has (finally) resolved this issue ... a few more Docker users need to weigh in with their results, but it looks pretty promising.

 

A couple of other "nits"  (one release-related; one not) ...

 

(1)  I've noticed this for a while, but assumed it would be fixed before final ... which is now VERY close, so thought I'd at least mention it => The Main page shows 25 slots available regardless of whether you boot to Pro or Plus.    (Don't have a Basic key to test with, but I suspect it may show them there as well)    Shouldn't it only show the # of slots actually available for your key ??

 

(2)  I mentioned it to Jon last week -- clearly it's not release-related and likely got "buried" in the "ought to do's" ... but it's also a trivial thing to fix, and I hate to see technical errors "glaring at you" on your web page ==>  the specifications for your AVS 10/4 are incorrect ... they show the CPU is a "Intel Xeon E3-1231V3 Ivy Bridge"  processor.  The 1231v3 is NOT an Ivy Bridge CPU ... it's a Haswell 

 

Hopefully this has (finally) resolved this issue ... a few more Docker users need to weigh in with their results, but it looks pretty promising.

 

A couple of other "nits"  (one release-related; one not) ...

 

I've noticed this for a while, but assumed it would be fixed before final ... which is now VERY close, so thought I'd at least mention it => The Main page shows 25 slots available regardless of whether you boot to Pro or Plus.    (Don't have a Basic key to test with, but I suspect it may show them there as well)    Shouldn't it only show the # of slots actually available for your key ??

That's hard because there are both disk slots and pool slots.  Say you have a Basic key: you can have 1 array slot and 5 pool slots or vice-versa.  I suppose we could have set them both to 6 but then there's be 12 slots displayed.  A couple releases ago I did try to change the default to something like 6 slots (can't remember exact number at the moment) but this caused some other issues.  Maybe we'll address this in the future.  Anyway the number of slots is independent of the number of devices.

 

I mentioned it to Jon last week -- clearly it's not release-related and likely got "buried" in the "ought to do's" ... but it's also a trivial thing to fix, and I hate to see technical errors "glaring at you" on your web page ==>  the specifications for your AVS 10/4 are incorrect ... they show the CPU is a "Intel Xeon E3-1231V3 Ivy Bridge"  processor.  The 1231v3 is NOT an Ivy Bridge CPU ... it's a Haswell 

Huh.  Nice find.

(1)  I've noticed this for a while, but assumed it would be fixed before final ... which is now VERY close, so thought I'd at least mention it => The Main page shows 25 slots available regardless of whether you boot to Pro or Plus.    (Don't have a Basic key to test with, but I suspect it may show them there as well)    Shouldn't it only show the # of slots actually available for your key ??

What if you have a Pro configuration on a new flash but don't have a Pro key for it yet? In that case it should show all your drives but just not let you start, so maybe it makes sense to show all slots.

a few more Docker users need to weigh in with their results

 

 

perhaps a post about it in the docker board ?

... That's hard because there are both disk slots and pool slots.  Say you have a Basic key: you can have 1 array slot and 5 pool slots or vice-versa.  I suppose we could have set them both to 6 but then there's be 12 slots displayed.  A couple releases ago I did try to change the default to something like 6 slots (can't remember exact number at the moment) but this caused some other issues.  Maybe we'll address this in the future.  Anyway the number of slots is independent of the number of devices.

 

I'd have thought the total # of slots would be limited to the # of assignable devices.  It actually works that way, but it always totals the 25 for a Pro key.  i.e. if you drop the # of drive slots then you can increase the # of cache slots.  But the total you can set for the two combined never exceeds 25.  In other words, it works exactly as I'd expect it to work for Pro ... so clearly the logic is there to ensure the total doesn't exceed the appropriate max (of 25 in this case).      On the surface, it seems like the "max" simply needs to be set to the appropriate # for the key.

 

 

1st reboot and every one of my containers now has internet connectivity.

 

+1

 

Thanks for squashing that pesky bug!

Sounds like you've cracked it, at work and actually having to do some tonight, will check in morning and confirm it's all working and post back.

 

Well Done LimeTech!

 

Knew you could do it.

 

 

 

kind of , my dockers work but now i have nothing to bitch about, lol.

 

Sure you have.... Two words......

 

Myth TV

 

kind of , my dockers work but now i have nothing to bitch about, lol.

 

Sure you have.... Two words......

 

Myth TV

 

... can't seem to find where LimeTech has advertised that as a built-in feature  :)

(thus NOT a valid thing to complain about)

 

... can't seem to find where LimeTech has advertised that as a built-in feature  :)

(thus NOT a valid thing to complain about)

 

You should see some of the PMs he's sent me about it, Sparkly clearly didn't get your memo.  Even I told him to have a rest and do something else! Lol

... can't seem to find where LimeTech has advertised that as a built-in feature  :)

(thus NOT a valid thing to complain about)

 

You should see some of the PMs he's sent me about it, Sparkly clearly didn't get your memo.  Even I told him to have a rest and do something else! Lol

So I'm not the only one he complains to then... 

 

But seriously Tom, Eric (and everyone else who was testing out the interim fixes), thank you for all your extremely hard work over the past couple of days.

So I'm not the only one he complains to then... 

 

No, in fact, I think he must be my online wife...

1. Moans

2. Gets tons done and is way more productive than I am

3. Craves attention

He does however have a better beard than Mrs CHBMB

I have just updated my Backup Server from 6.0-rc4 to 6.0-rc6a via the web-gui Plugin Tab.

 

Upgrade seemed to work flawlessly. The server started as did the array without an issue I could see.

 

A quick cursory look around sees nothing amiss. I don't run any Dockers on my Backup Server but I do run a Windows 10 VM and it started without issue.

 

Clearly there has been some good work done around the web-gui (system info, disk shares etc) and I like them all so far.

 

This method of updating Unraid OS is a great great feature.

 

Pat's on the back for LT I think. Looking good.

... This method of updating Unraid OS is a great great feature.

 

Definitely agree ... my last couple upgrades have been with this nifty little button instead of the old download/extract/copy-to-the-flash-drive/logon to server/reboot approach.    MUCH nicer and certainly more user friendly 

Had a more detailed response typed up but then my browser gave up the ghost and crashed, so apologies for the next set of run-on phrases...

 

Under RC6 I edited docker.cfg to remove the --dns setting, upgraded via the UI to RC6a, stopped the array, rebooted, waited for the system to come back up, dockers started up fine and functioned despite not having completed the full upgrade instructions.

 

In examining them, I noticed the /etc/resolv.conf didn't exactly match my host file, even though the settings were identical, so the OCDp kicked in. I then removed all docker containers and images, stopped docker system, deleted the docker.img file, started docker system, and verified it recreated the docker.img. I then re-added the first docker container, watched as it pulled down the images, started the container, noted everything worked, examined the docker /etc/resolv.conf and noted it now matches exactly the host's file. I then re-added the second and third docker containers and noted they work exactly as expected.

 

Finally, I set all the dockers to auto-start.

 

I'm also liking the docker image usage is back down to it's compact size of 617 Megs. All the editing and recreating of the images had ballooned the image usage up to 2.4GB - 3.2GB.

 

Who would have thought there would have been larger implications relating to a bug reported in November/December 2014? Good job on the timely release!

 

That's interesting BRiT, not sure I completely understand the implications though.

 

If we deleted and recreated oir docker.img and redownloaded all our containers, would that have been a fix for the DNS bug? Was going to be my next step if RC6a hadn't fixed stuff.

 

I didn't realise the docker image would grow! Doesn't it purge data that is no longer required?

All appears well with rc6a - many thanks to all.