unRAID Server Release 6.2 Stable Release Available


limetech

Recommended Posts

I occasionally get errors when hard linking still. I've been using it ever since it was available in the RC. I am still on RC4 as I haven't had a chance to shut down my VMs and reboot but I don't believe there is a difference between that and the latest release as far as hard linking goes. I can troubleshoot more once I get the chance but I am getting "Invalid cross-device link" when trying to create a hard link between shares. My disk are getting pretty full so that might be relevant. Any ideas?

You can't create a hard link across shares.

Link to comment
  • Replies 443
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Posted Images

little glitch after upgrading to 6.2

have array set to auto start but it does not after a reboot.

also not capturing current log to flash drive on shutdown. attached diagnostics run from tool menu after starting array.

I had the depreciated powerdown plugin installed on 6.1.9 which i uninstalled prior to upgrading to 6.2 if that means anything useful.

i also use the dnyamix button from the dashboard to initiate the reboot.

If i've done something, help shed light on it so i can reverse it. thanks.

tower-diagnostics-20160923-1640.zip

Link to comment

Any reason why I cannot get tabbed mode in display settings?  I change it back to tabbed, and when I apply it, it changes back to non-tabbed.

 

No reason. Have you tried different browsers?

 

Might be worth to check the flash device too and see if it saves the updated settings.

 

Tried several browsers.

 

Plus, my system has been set for tabbed interface for a long time.  Upgrade to 6.2 stable changed it to untabbed.

Link to comment

Any reason why I cannot get tabbed mode in display settings?  I change it back to tabbed, and when I apply it, it changes back to non-tabbed.

 

No reason. Have you tried different browsers?

 

Might be worth to check the flash device too and see if it saves the updated settings.

 

Tried several browsers.

 

Plus, my system has been set for tabbed interface for a long time.  Upgrade to 6.2 stable changed it to untabbed.

Can you still write to the flash drive?

 

Sent from my SM-T560NU using Tapatalk

 

 

Link to comment

- add experimental global share setting: "Tunable (enable Direct IO)"

 

Update on this setting, I've been using it since rc5 was release without any issues, array was up for 5 days, but some time a go (maybe 1 or 2 days) speed reverted back to what I used to get with the setting off, <800MB/s, stopping and starting the array fixed and speed is >1GB/s again, will test every day now to check if it happens again.

 

After some more days testing can confirm there's an issue here, using Direct IO, transfers to user shares are done at full 10GbE speed after array start, after a few hours speed degrades to between 800MB/s and 950MB/s, to regain full speed I need to stop/start array.

 

It's not my system/network as I always get full speed transferring to a disk share, even to unassigned devices.

 

It's also not a big deal and speed is close enough (and higher than with Direct IO off), just though I should report it.

Link to comment

Here is my experience with 6.2.  Suffice to say it was a failure.

 

I logged in to unRAID today to see that there was an upgrade listed in the Plugins menu.  I am running 6.1.9 so figured why not give a try. 

 

Here are the steps I followed.

 

1. Updated all other Plugin first. (Powerdown Package, FCP, Dynamix System Plugins)

2. Stopped all Dockers

3. Backed up entire Flash contents.

4. Updated unRAID via the button in Plugins

5. Rebooted server per instructions on Plugins page

6. The boot up took about 9 mins but finally booted to normal terminal prompt (non gui)

7. Shares did not work, couldn't access unMenu, could log in though so did that but couldn't do anything else.  Ran ifconfig and it reported my ethernet was missing.

8. Rebooted in to new Gui boot up method.  Took about 3 min to get to Gui but a full 10 min to finally show the unRAID interface page.

9. I have two NIC's so went to settings and found that 6.2 switched my NIC assignents so eth0 (the adapter that was configured for unRAID) was the unused NIC

10. Swapped the assignments so the proper NICs were in use.

11. Rebooted per the instructions back in to GUI mode

12. Booted up to firefox in about 5 min however too another 10 minutes for the bootGUI webmenu to finally load

 

Once booted up things began to quickly fall apart.  The GUI screen was not working right.  It was taking forever to load and switch tabs.  Images would not load for the Dashboard tab.  The shares were not accessible.  My Dockers apps were not reachable.  All Docker apps showed an unknown version in red.  My main VM wouldn't load, just threw up errors.

 

I tried to fool around with stuff and search online for answers but I couldn't find anything so I deleted everything and restored the backup.  I am now back running on 6.1.9 and enduring a parity check because of an unclean shutdown.

 

If limetech would like to use me as a test bed for an updgrade to help out others then please let me know but right now it appears that a clean upgrade is not in the cards for a complex 6.1.9 setup

Link to comment

Probably a big chunk of the time delays were in docker converting the image file to be compatible with 6.2.  Documented in the OP and a one-time procedure.

 

The docker apps being listed as unknown version is a predecessor warning to the image layers do not match manifest issue (ultimately a dockerHub issue, not a unRaid issue) which is well documented in the OP

 

Without diagnostics however from 6.2, its only a guessing game as to what actually happened however.

 

Link to comment

7. Shares did not work, couldn't access unMenu, could log in though so did that but couldn't do anything else.  Ran ifconfig and it reported my ethernet was missing.

 

Sounds like you have some old and incompatible extensions installed. Get rid of unMenu and clean up your go file. Remove plugins then reinstall them one at a time.

 

Link to comment

I upgraded to 6.2 yesterday and several things did not work (and are listed as potentially not working). My docker tab is missing but what is confusing for me is that in reading the help around this and docker paths, all seems fine. My image was in /mnt/cache/ and my appdata is in /mnt/user/appdata/.

 

I do see the above may be caused by my docker.img being on an "unassigned drive" but I do not really know what this means? It is great to have these tips but I could use some more details on what some of that means in the notes.

 

My docker status - in the docker setting tool is STOPPED. I think on one reboot, it was listed as SUSPENDED. Obviously this is an issue. How do start the docker service? I see no way to do this in the GUI.

 

I am also seeing this error if I try to reinstall a docker:

 

Warning: stream_socket_client(): unable to connect to unix:///var/run/docker.sock (No such file or directory) in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 671 Couldn't create socket: [2] No such file or directory Warning: Invalid argument supplied for foreach() in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 935

 

Warning: stream_socket_client(): unable to connect to unix:///var/run/docker.sock (No such file or directory) in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 671 Couldn't create socket: [2] No such file or directory Warning: stream_socket_client(): unable to connect to unix:///var/run/docker.sock (No such file or directory) in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 671 Couldn't create socket: [2] No such file or directory Warning: Invalid argument supplied for foreach() in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 851

 

I am tempted just to remove my docker.img and start again. It is pretty easy to reinstall dockers these days.

 

Any suggestions on the above before I either blow away my docker.img or roll back to 2.1.9?

 

Log attached from after I reboot. There is clearly something wrong as my docker image is "in use" and cannot be mounted.

 

Thanks!

log.txt

syslog.txt

Link to comment

I upgraded to 6.2 yesterday and several things did not work (and are listed as potentially not working). My docker tab is missing but what is confusing for me is that in reading the help around this and docker paths, all seems fine. My image was in /mnt/cache/ and my appdata is in /mnt/user/appdata/.

 

I do see the above may be caused by my docker.img being on an "unassigned drive" but I do not really know what this means? It is great to have these tips but I could use some more details on what some of that means in the notes.

 

My docker status - in the docker setting tool is STOPPED. I think on one reboot, it was listed as SUSPENDED. Obviously this is an issue. How do start the docker service? I see no way to do this in the GUI.

 

I am also seeing this error if I try to reinstall a docker:

 

Warning: stream_socket_client(): unable to connect to unix:///var/run/docker.sock (No such file or directory) in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 671 Couldn't create socket: [2] No such file or directory Warning: Invalid argument supplied for foreach() in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 935

 

Warning: stream_socket_client(): unable to connect to unix:///var/run/docker.sock (No such file or directory) in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 671 Couldn't create socket: [2] No such file or directory Warning: stream_socket_client(): unable to connect to unix:///var/run/docker.sock (No such file or directory) in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 671 Couldn't create socket: [2] No such file or directory Warning: Invalid argument supplied for foreach() in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 851

 

I am tempted just to remove my docker.img and start again. It is pretty easy to reinstall dockers these days.

 

Any suggestions on the above before I either blow away my docker.img or roll back to 2.1.9?

 

Log attached from after I reboot. There is clearly something wrong as my docker image is "in use" and cannot be mounted.

 

Thanks!

 

Coming from 6.1.9 you need to delete the docker.img anyway, so just delete it and set it up again.

 

Link to comment

I think that either this should be automated, or made super clear as a message or something upon upgrading.

 

I actually read the thread, and didn't notice that the User has to delete the file.  Most people (rightly) assume that if an update is available, they can just click the button.

 

 

I think LT would be wise to improve the update process to handle such exceptions;  and the ones that can't be should be popped to the user in the GUI. 

 

This becomes more and more important as updates happen on updates, and potentially each needing independent "action".

 

You can't expect the Users going from (for example) 6.1 to 6.8 to go through every update thread to make sure they see every upgrade minutiae.

Link to comment

I think that either this should be automated, or made super clear as a message or something upon upgrading.

 

I actually read the thread, and didn't notice that the User has to delete the file.  Most people (rightly) assume that if an update is available, they can just click the button.

 

 

I think LT would be wise to improve the update process to handle such exceptions;  and the ones that can't be should be popped to the user in the GUI. 

 

This becomes more and more important as updates happen on updates, and potentially each needing independent "action".

 

You can't expect the Users going from (for example) 6.1 to 6.8 to go through every update thread to make sure they see every upgrade minutiae.

 

This is a great point.  I didn't even read the OP because there was an update button.  I just clicked update and "assumed", quite wrongly it seems, that unRAID would walk me through the update process.

Link to comment

 

As folks say, I did try to do some reading but assumed (like updates before 6.2) that system updates did not need "prep" work.

 

I would have happily just deleted my docker.img as it is super easy to reinstall dockers now. It was not (at least for me) when i first started.

 

I understand that there are many, many users and not all combinations can be tested but if there are some "good practices" to do before updating to 6.2 I think they should be called out on the update page? Just a thought,

 

That said, really excited to have 6.2 and give it a good run. Deleting my docker.img sorted it all out.

I got very confused as the suggested fix on the first page discussed the issue being an "unassigned drive" which I spent some time trying to figure out what that meant. I wanted to jump right in and delete the docker.img but was nervous.

Link to comment

little glitch after upgrading to 6.2

have array set to auto start but it does not after a reboot.

also not capturing current log to flash drive on shutdown. attached diagnostics run from tool menu after starting array.

I had the depreciated powerdown plugin installed on 6.1.9 which i uninstalled prior to upgrading to 6.2 if that means anything useful.

i also use the dnyamix button from the dashboard to initiate the reboot.

If i've done something, help shed light on it so i can reverse it. thanks.

 

update: still experiencing this situation.

i've found that if i leave the browser window open until the rebooting webpage is displayed all boot up fine.

if i close the browser before this page is displayed the situation occurs.

consistent for me with FF and chrome. Any ideas?

Link to comment

Here is my experience with 6.2.  Suffice to say it was a failure.

...

Ran ifconfig and it reported my ethernet was missing.

...

9. I have two NIC's so went to settings and found that 6.2 switched my NIC assignents so eth0 (the adapter that was configured for unRAID) was the unused NIC

10. Swapped the assignments so the proper NICs were in use.

 

I wanted to comment on a number of things, but I'll start with the networking.

 

I've added a note to the network section of the Additional Upgrade Advice about this, as it seems to happen to a few users almost every release where there's a significant upgrade of the kernel.  It's not the fault of 6.2, and nothing 6.2 could do about it, but it's certainly worth a note, something I've forgotten to add on this release.  I've added a note about it to previous Upgrade guides, but forgot it here.  It usually only affects a few users.

 

Here's the note I added:

- Whenever there is a significant upgrade of the underlying kernel, inevitably there are a few users with multiple NIC's that lose connectivity, because the kernel will identify the NIC's in a different order than previous kernels did.  That means eth0 will shift to a different NIC, perhaps one you aren't even using.  That also means your server may be offline, unreachable!  Often the fix can be as simple as switching the network cable to the other port, but usually you will want to study the situation, and determine what needs to be switched or reconfigured, to restore things the way you wanted them.  There are significant changes mentioned in the following that will impact this, going forward, and make it less of a problem, since you will be able to configure which physical port is eth0, eth1, etc, and expect them to stay that way even if a future kernel again changes its mind as to which to setup first.

And then following this, in the Additional Upgrade Advice (network section), is an introduction to all of the new changes in the network configuration, and the new network features.

 

I hope this will help at least with that aspect of the upgrade.  It seems like such a small thing, when the kernel identifies NIC's and associates them with networking symbols, but it can cause so much grief, when you can't even reach your server, especially if it's headless.

Link to comment

Just wanted to post a positive experience. Upgraded from 6.1.9. I removed the power down plugin before doing the update. Everything updated from the gui without a hitch!

 

I removed a small drive and replaced it with a larger drive for a 2nd parity and that went great as well. I've only update on docker container but it updated fine and started up without problems. I already had an appdata folder on my cache drive so I'm guessing that had something to do with some of the success.

 

Thanks for all of everyone's hard work.

Link to comment

I think that either this should be automated, or made super clear as a message or something upon upgrading.

 

I actually read the thread, and didn't notice that the User has to delete the file.  Most people (rightly) assume that if an update is available, they can just click the button.

 

I think LT would be wise to improve the update process to handle such exceptions;  and the ones that can't be should be popped to the user in the GUI. 

 

This becomes more and more important as updates happen on updates, and potentially each needing independent "action".

 

You can't expect the Users going from (for example) 6.1 to 6.8 to go through every update thread to make sure they see every upgrade minutiae.

 

This is a great point.  I didn't even read the OP because there was an update button.  I just clicked update and "assumed", quite wrongly it seems, that unRAID would walk me through the update process.

 

As folks say, I did try to do some reading but assumed (like updates before 6.2) that system updates did not need "prep" work.

 

I would have happily just deleted my docker.img as it is super easy to reinstall dockers now. It was not (at least for me) when i first started.

 

I understand that there are many, many users and not all combinations can be tested but if there are some "good practices" to do before updating to 6.2 I think they should be called out on the update page? Just a thought,

 

That said, really excited to have 6.2 and give it a good run. Deleting my docker.img sorted it all out.

I got very confused as the suggested fix on the first page discussed the issue being an "unassigned drive" which I spent some time trying to figure out what that meant. I wanted to jump right in and delete the docker.img but was nervous.

 

I wanted to comment on certain misconceptions here, as well as on the implied advice for future releases -

 

* It is absolutely imperative to read the announcement posts.  It's also good to read any additional upgrade advice that may be available.  In the past, we didn't have the built-in plugin upgrade tool, so all users learned of upgrades through the forum and generally read the announcements, and learned of extra steps necessary, and gotchas to avoid.  Which is why it has probably never occurred to anyone before now, that some users might only learn of an upgrade through the program and then never read the announcement.  That's a good point, something to recommend to Tom, and I suspect he is already planning on ways to make sure that upgrade announcements in the program include strong recommendations to read the announcement post.

 

* This is not an app being updated, this is an OS.  Can you imagine upgrading from Windows 8 to Windows 10 without doing a lot of reading?  You need to know about the new features, the changes in the way things operate, the things that need to be converted or replaced, the decisions to make, etc.  I'm positive the extra steps in the 6.2 upgrade were not chosen lightly, but were found to be required in order to use newer Docker and VM technologies.  And there are probably more of those changes coming, as both are still a moving target.

 

* Comments were made to the effect "why weren't we notified of certain steps".  I don't think most of them were known before very recently, or they would have been mentioned.  I see that now there's a recommendation to just delete docker.img, when coming from 6.1.9, and I'll add a note to the Additional Upgrade Advice about that.  But that is a new idea, only within the last day or 2, and I believe many did not and do not have to do that, and had no issues.  You naturally only hear from those with Docker issues here.  It does seem like good advice, for anyone coming from 6.1.9 *and* having Docker issues (like no Docker tab or none of the Dockers working or "layers" error or all need updates, etc).

 

* As to the confusion about "unassigned drive", that was my fault in not making it clear and specific.  I was trying to stay generic, but it is referring to storing your docker.img and/or appdata folder on a drive mounted through the Unassigned Devices plugin.  I will try to clarify that.

 

* I do believe the original announcement does mention everything that was known about the upgrade at the time it was written.  As other issues and clarifications have arisen, we've been trying to include them in the subsequent post - Additional Upgrade Advice, I recommend reading through it.  It's still changing, more items are still being added.

 

Just to be sure it's clear, I'm a fellow user too, just trying to be helpful.  I don't work for Limetech, so nothing I say is official!

Link to comment

* This is not an app being updated, this is an OS.  Can you imagine upgrading from Windows 8 to Windows 10 without doing a lot of reading?  You need to know about the new features, the changes in the way things operate, the things that need to be converted or replaced, the decisions to make, etc.  I'm positive the extra steps in the 6.2 upgrade were not chosen lightly, but were found to be required in order to use newer Docker and VM technologies.  And there are probably more of those changes coming, as both are still a moving target.

 

Uhh thousands of people upgrade Windows/OSX probably daily without reading anything about it. Those upgrades handle incompatibility and let the user know about it. I get what you are saying but that was a terrible analogy.

 

Since we now have a button that upgrades in-app I don't think its fair to assume that means people should still know to go read and announcement thread. We have been conditioned that both app and OS updates are virtually seamless for everything nowadays, Windows, OSX, Android, iOS and all the apps within them.

Link to comment

I believe some of this could be prevented if LimeTech would consider adding a couple of items. 

 

First, on the Plugin  page when a upgraded version of the Server OS become available, modify that block with a link to the update announcement thread and a statement to say 'read before updating'! 

 

Second in the update announcement, change the first line from:

 

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

 

to something like this:

 

Clicking 'Check for Updates' on the Plugins page is the preferred way to upgrade.  But before doing the update, carefully read 
the first/two posts for additional information which may affect your particular situation.  

 

 

Link to comment

* This is not an app being updated, this is an OS.  Can you imagine upgrading from Windows 8 to Windows 10 without doing a lot of reading?  You need to know about the new features, the changes in the way things operate, the things that need to be converted or replaced, the decisions to make, etc.  I'm positive the extra steps in the 6.2 upgrade were not chosen lightly, but were found to be required in order to use newer Docker and VM technologies.  And there are probably more of those changes coming, as both are still a moving target.

 

Uhh thousands of people upgrade Windows/OSX probably daily without reading anything about it. Those upgrades handle incompatibility and let the user know about it. I get what you are saying but that was a terrible analogy.

 

Since we now have a button that upgrades in-app I don't think its fair to assume that means people should still know to go read and announcement thread. We have been conditioned that both app and OS updates are virtually seamless for everything nowadays, Windows, OSX, Android, iOS and all the apps within them.

 

This is a fair point. I do not remember being told by the web gui that I "now need to go other stuff manually".

Link to comment

I think that one way to sort this issue (and might be a good generic option for plugins anyway) is to have the release notes displayed automatically when you select to install the update, and the install only actually happens when you click a button that is at the end of these notes.  I would think that it would be acceptable to update such notes with new information that is discovered so that you always get the latest version of such notes.

 

If this was done well then initially the note would be a duplicate of what is in the forum posting, but it would be updated as time goes by with further hints on updating in resolving any issues around the install that are discovered from early users experiences.  If one did not want the effort of maintaining such notes separately to forum updates then this could be handled by providing links to the relevant forum posts.

 

Not sure what happens about auto-updating plugins without user intervention, but I am sure that a little thought could resolve this in a satisfactory issue as well.

Link to comment

- add experimental global share setting: "Tunable (enable Direct IO)"

 

Update on this setting, I've been using it since rc5 was release without any issues, array was up for 5 days, but some time a go (maybe 1 or 2 days) speed reverted back to what I used to get with the setting off, <800MB/s, stopping and starting the array fixed and speed is >1GB/s again, will test every day now to check if it happens again.

 

After some more days testing can confirm there's an issue here, using Direct IO, transfers to user shares are done at full 10GbE speed after array start, after a few hours speed degrades to between 800MB/s and 950MB/s, to regain full speed I need to stop/start array.

 

It's not my system/network as I always get full speed transferring to a disk share, even to unassigned devices.

 

It's also not a big deal and speed is close enough (and higher than with Direct IO off), just though I should report it.

 

Update on this issue, should be the last one, I hope  :P

 

I upgraded my server on the weekend and since then I'm not seeing this issue anymore, speed writing to user shares remains consistently high when using direct IO, also max speed is a little higher than before.

 

So, though it wasn't just my system, as I could get full speed writing to a disk share, it does look like the system used has some effect when writing to user shares, it's a little surprising to me since the new one is not much better than the old, from Supermicro X9SCL-F with Xeon E3-1220 to Supermicro X11SSM-F with Xeon E3-1220v5.

Screenshot_2016-09-26_11_59_09.png.b6566499108430a575ab95baf36994e6.png

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.