5 hours ago, ich777 said:

Please don‘t double posts!!!

I‘m already responded in the ZFS Support Thread.

 

My apologies for the the double post, and thank you for replying.  I have deleted my post above. 

19 hours ago, limetech said:

This release is focused on bug fixes and minor improvements.

 

Some of us over in this thread would appreciate some of this 6.11 series bug squashing love:

 

-JesterEE

I upgraded from 6.11.1 to 6.11.3 and as soon as I rebooted i can no longer boot, as I am getting Boot Error

15 minutes ago, dorgan said:

I upgraded from 6.11.1 to 6.11.3 and as soon as I rebooted i can no longer boot, as I am getting Boot Error

Switching to EUFI boot made me be able to boot again, so I am not sure whats going on.

Upgrade went smoothly. Boot took just over 7 minutes, hanging with the last output being Samba stuff for almost 5 minutes.

Same bug as 6.11.2 where I had to tick "show password" to get the "Start" button enabled for the array (password filled via password manager, not typing or copy/paste)

upgraded from 6.11.2 with no issues.  Thanks for making it smooth

10 hours ago, craigr said:

Yes, my pool devices were assigned again (thank the maker).  I did of course have a backup from 6.11.1.  I recreated the flash drive from scratch using a fresh download of 6.11.1 and copied over my config directory.

 

 

Funny thing too is that my SSD cache pools use the SATA ports on the MB and the cache drive uses a port on my LSI controller.  Why lose ALL cache drives?!?

 

10 hours ago, craigr said:

I can try 6.11.3 again later when I have more time and get the log files.  Right now, I just need a functioning computer.

 

10 hours ago, JorgeB said:

Please do, and if you see the same issues post the diags in a new thread in the general support forum.

 

Well, I updated from 6.11.1 to 6.11.3 and the problem did not reoccur.  My cache disks are all assigned and intact.  I have rebooted multiple times to test and it's all fine.  Something with the upgrade between 6.11.2 to 6.11.3 obviously went sideways that did not happen after I reverted back to 6.11.1 and upgraded from there.

 

Oh well.

 

Now if I could only get my Windows 11 VM to boot more than once with boot order enabled...  but I have a work around for that so I can live with it for now.

 

Thanks,

craigr

I just logged onto the forum to get some help with a 12TB drive I was adding to my array that wasn't formatting after pre-clear... If only that upgrade notification came earlier! Will report back if 6.11.3 doesn't solve the problem...

 

Update - Formatting working as expected now. Thanks!

Edited November 10, 20223 yr by sreknob
update

Just made an upgrade from 6.10.3 to 6.11.3. No issues, thanks.

I normally don't post the gory details of a bug and bug fix/workaround.  For the issue where 6.11.2 could not format devices larger than 2TB here is what was the cause and the fix that went into 6.11.3.

 

First, the reason this got published without this issue being caught is that my test server with devices larger then 2TB were already formatted.  Other test servers, which are actually VM's had smaller virtual devices and formatting worked ok with them.  So sorry about that, sometimes sh*t happens, but this particular bug would not have caused any data loss.  Anyway....

 

The bug was in the 'sgdisk' utility included in the "gptfdisk" package, where version 1.0.9 was released on April 15, 2022.  This was integrated into Unraid OS very early in 6.11 development and appeared in 6.11.0 on Sep 23, 2022, where it works fine.  But then the libpopt library was updated on Oct 19, 2022, and included in a larger slackware package called "aaa_libraries" which was integrated into Unraid 6.11.2.  The libpopt library includes functions used by nearly all C-based utilities to parse command line options.

 

The sgdisk command also uses this library to parse command line options. The bug is that sgdisk was referencing freed memory due to how a function in libpopt was implemented. But this didn't cause problems because that freed memory wasn't reused until after sgdisk already referenced it. Well along comes an upgrade to the libpopt library which frees memory differently which caused sgdisk now to reference garbage data and fail as described.

 

Author of gptfdisk added a fix but for some reason has not published a new release.

 

Edit: the fix was to revert gptfdisk to 1.0.8 which used the earlier version of libpopt, and to also revert libpopt to the earlier version.

Thanks for the transparency.   Also, a big thank you for identifying and fixing the issue rather quickly.

 

Although I was not impacted by the bug, many new users or existing users building new systems or upgrading disks were impacted.  It is easy to see how this particular issue fell through the cracks and went undetected until release; however, as it is now a known scenario hopefully it has been added to future pre-release testing regimens.  It's kind of a big deal when a disk cannot be added to an array because it cannot be formatted.  😁

Upgraded to 6.11.3 and it went well but now I lost all of the Shares.  A reboot fixed it but this worries me.

 

Also, my Windows 11 vm with gpu passthrough is working as normal but then suddenly crashes and reboots.  I had to re-create the VM last time.

 

So few funky things happening.

I not a linux expert so I don't know how to fix this. I'm able to log onto the system locally but I cannot get to the system remotely.

1 minute ago, kevon27 said:

get to the system remotely

My Servers plugin and/or Wireguard VPN

Just now, trurl said:

My Servers plugin

https://forums.unraid.net/forum/94-my-servers-plugin-support/

2 minutes ago, trurl said:

Wireguard VPN

 

Let me rephrase my statement. After the update to 6.11.3, my server is  no longer reachable on my network. I cannot ping or ssh into the machine.

5 minutes ago, kevon27 said:

my server is  no longer reachable on my network. I cannot ping or ssh into the machine.

Sometimes, the network.cfg and network-rules.cfg can get messed up.  It has been known to happen in upgrades.  Delete these files from the config folder on your unRAID flash drive and reboot.  New default files will be created but your server will get a DHCP address.  You can then change network settings if you need to.

 

Alternatively, if you have a flash drive backup from before the problem occurred, you can try restoring these files from the backup.

Edited November 10, 20223 yr by Hoopster

Ok.. I fixed the problem. My MTU settings was set to high - 9000 (this is what happens when you like to tinker with stuff 😁). I edited the network config and set MTU back to 1500.

Unraid is back online.

Anyone having issues with virtiofs and windows? I heard older versions were having issues and I'm wondering if they're still happening.

4 minutes ago, tech3475 said:

Anyone having issues with virtiofs and windows? I heard older versions were having issues and I'm wondering if they're still happening.

There is nothing changed for the memory backing issue. But it doesnt affect alll setups.

I just updated from 6.10.3 and all my VM VNC stopped working?

 

I see some options have changed in the settings but I can't figure out why it will not connect? Kinda need one of them quickly.

Just tried installed a fresh VM and the VNC is not working on it either, was something broken with this update? Do I need to revert?

Just now, TexasUnraid said:

VNC is not working on it either

Clear browser cookies.

1 minute ago, JorgeB said:

Clear browser cookies.

Thanks! Didn't even cross my mind. Sure enough that did it, now to go reset all my settings for my dockers that use cookies lol

 

While I was trying to figure this out I noticed that there is the option for both VNC and SPICE in the VM? What is the difference? Is one better then the other?