11.2 - Who has installed?

[T.W]

Upgraded 3 machines from 11.1-U6 to 11.2 the only issue that I can see so far is the status of cloud sync tasks - it's a constant revolving circle

 

[Spud]

A lot of graphs not working properly just reading zero, I'm switching back to the legacy UI until that gets sorted out, hopefully it won't be long.

Plugins and VMs seem to be working fine.

Need to turn this email notification thing off though its driving me crazy

 

[KrisBee]

Right, I have checked the RancherOS and only if you create a new VM, it is created as v1.4.2. Existing RancherOS VM is still at v1.4.1. Is there a way to upgrade existing RancherOS instance to the latest ?

IIRC, it was possible in the past. But the sequence was to upgrade the rancheros first via ros commands, and then upgrade FreeNAS itself. Personally, I'd want to test this very carefully before trying to upgrade an existing v1.4.1 "docker vm". There may be a way to hack this at the CLI, but again you'd have to proceed with caution.

 

[dan3408]

Updated yesterday from 11.1u6, and things went great. Surprisingly easy to migrate over my old jails and plugins to iocage.

I have only a couple issues I am still noticing:
* https://redmine.ixsystems.com/issues/62583 Disk temp not shown
* my first jail won't open a shell from the UI, https://forums.freenas.org/index.ph...rom-gui-and-only-shows-login-incorrect.71772/

Thanks for a great release and keep up the good work!

 

[ThreeDee]

hmm ... just tried 11.1-U6 to 11.2 and failed to boot "error 2" .. had to go back to 11.1-u6 to get things to boot again. I'll try the update again and see what happens.

 

[Disruptor]

I updated my NAS from 11.1u6, the update failed and was throwing quite a lot of error messages during boot/migration (something about host unreachable, client timeout and a couple stack traces but it passed so quickly that I couldn't read them). After a couple minutes it showed the regular login screen on the console, but the web interface was only showing a message that my nas seemed to be offline and I should make sure it was running... I was able to resolve the issue following: https://forums.freenas.org/index.php?threads/11-2-beta-1-login-issues.68448/page-2#post-478414

Code:

$ cat /data/update.failed
CommandError: Conflicting migrations detected; multiple leaf nodes in the migration graph: (0018_auto_20180918_1848, 0020_make_lunid_non_null in services; 0005_merge_20180617_1001, 0005_auto_20180918_1848 in tasks; 0013_remove_cifs_default_permissions, 0012_auto_20180918_1848 in sharing; 0028_cert_serials, 0023_auto_20180918_1848 in system).
To fix them run 'python manage.py makemigrations --merge'

Seems like something went wrong during the migration but the fix using manage.py mentioned in the post did solve it. The new web interface doesn't show statistics for me in the dashboard (they work in the reports section), but I use an external influxdb+grafana server which still works, so that's fine for me. Also one of my pools got stripped of the encryption password, and another did not? How ever I was able to reset the password and rekey them without issue.

These issues aside I'm positively surprised by the new UI, it looks nice and was working great for the things I did test wise. Also the general load seems to be better and the arc hit ratio went up from 12% to 99% (need to do more tests and see how it performs with the day to day load)


Thank you everyone for their hard work on this release!

 

[danb35]

Also one of my pools got stripped of the encryption password

Another user mentioned this up-thread as well--I'd suggest filing a bug from the system where that happened, and attaching a debug file, so hopefully the devs can figure it out.

 

[diskutor45]

Is it safe to upgrade directly from 11.1-u6 to this with plex?

I have made an upgrade from 11.1 U6 to 11.2 yesterday. Upgrade process was fast and without issues everything was working including Plex (Im PlexPass subscriber).
However Plex was running in the "old" jail accessible by legacy UI.
I have decided to migrate it to iocage by installing Plex Media Server plugin using the new UI, created mount points according to the original jail and then followed the manual:
https://support.plex.tv/articles/201370363-move-an-install-to-another-system/

After the migration I had to download updated script for Plex upgrades:
https://github.com/luluhoc/PMS_Updater

Currently its running fine but I will wait a while before I do the final step - upgrade of the ZFS pools.

Thank you FreeNAS developers for a good work. Its great.

 

[Futs]

I guess I wasn't clear. I was looking for a way to disable only email notifications. I do understand that disabling netdata completely would disable email notifications along with it. Before this update, I had never received any email notifications from netdata, so obviously something changed which caused me to get about 300 emails.

I'm not sure if this is the correct way to do it, but setting SEND_EMAIL="NO" in /usr/local/etc/netdata/health_alarm_notify.conf and restarting the service seems to have done what I was looking for.

I just changed DEFAULT_RECIPIENT_EMAIL="root|critical", will see how that works out as the majority of alerts have been "Needs Attention" or "Recovered"

 

[babaroga]

IIRC, it was possible in the past. But the sequence was to upgrade the rancheros first via ros commands, and then upgrade FreeNAS itself. Personally, I'd want to test this very carefully before trying to upgrade an existing v1.4.1 "docker vm". There may be a way to hack this at the CLI, but again you'd have to proceed with caution.

I did try "ros" upgrade commands while in RC2, upgrade never worked until I did RTFM. Still, now that release has 1.4.2, on every reboot, VM goes back to v1.4.1

Code:

root@radfree:~ # ssh rancher@192.168.1.200
rancher@192.168.1.200's password:
[rancher@rancher ~]$ sudo ros os version
v1.4.1
[rancher@rancher ~]$ sudo ros os list
rancher/os:v1.4.2 local latest
rancher/os:v1.4.1 remote available running
rancher/os:v1.4.0 remote available
rancher/os:v1.3.0 remote available
....

 

[felthamn]

I installed it last night. No issues with the upgrade. Jails, VM's seem to be working okay. Had a play about with the plugins and some seem a bit flakey. Could not install redmine just stuck at 75%. It's not really an issue to me as I prefer to create my own jails and install manually (thanks to the advice of others on this forum). The only other thing of note is the graphs aren't working well - see the attached file. Just for the record I don't see any smoke pouring out the back of the server. This was a pretty good upgrade with no hitches so far.

 

[KrisBee]

I did try "ros" upgrade commands while in RC2, upgrade never worked until I did RTFM. Still, now that release has 1.4.2, on every reboot, VM goes back to v1.4.1

Code:

root@radfree:~ # ssh rancher@192.168.1.200
rancher@192.168.1.200's password:
[rancher@rancher ~]$ sudo ros os version
v1.4.1
[rancher@rancher ~]$ sudo ros os list
rancher/os:v1.4.2 local latest
rancher/os:v1.4.1 remote available running
rancher/os:v1.4.0 remote available
rancher/os:v1.3.0 remote available
....

That's the expected behaviour as the boot config files in the hidden ".bhyve_containers" pool folder older still point to the 1.4.1. kernel and initrd. It's the equivalent of updating the kernel in linux, but then choosing to boot an older kernel from a grub menu when rebooting.

Probably better to start a separate thread about this under "Jails, Plugins and Bhyve".

 

[Maxburn]

Things went sort of well. I took the opportunity to move off of boot USB to a SSD. Heard many times that moving the install from one boot media to another was as simple as restoring the config, and while that is mostly true it doesn't restore SSH keys which are required for my remote backup tasks. So just spent some time re-keying everything. If I just did a straight upgrade that likely would have worked perfect.

New UI seems to work better in my chromebook. Old UI required page refreshes to get the pop up windows working again, so thats a win.

I'm also seeing the weird graph not filling in issue but seems to work if I refresh, or at least different ones don't work after a refresh.

My HDD temps seem to be working now, don't think that worked for me before. That's a big deal because I found out I had a fan unplugged and one was rather warm.

 

[InQuize]

11.1-U6 => 11.2-RELEASE
Upgraded without a hitch and everything seems working as expected (I only use warden jails right now; SMB, NFS, Smart, UPS all ok).

Show : 12.2-RELEASE comments

The only annoying thing that I noticed immediately is the notification about pool being used over 90% was sent to my e-mail, and I have 4 of them filled to the top (I know it is usually not recommended, but these pools are just separate single disks for non-critical data that do not use snapshoting or any other ZFS feature except compression, so it's totally fine and works here for years since 9.3).
If this notification is going to be sent every time percentage fluctuates, that would be very irritating and there should definitely be some option to disable that, ideally for each individual pool.

Show : New UI comments

Not a fan of new interface though. It feels "over simplistic", the old one was not crowded at all IMO, on the contrary it's layout was very logical and intuitive. I feel like use of space in new one is so poor that it makes it annoying, like it was designed for grandma. Especially new graphs look gigantic comparing to RRD. I feel like person that designed that GUI is a fan of "Metro UI" which is not very successful, at least IMO. All those tiles are meant for touchscreens to fit under one's fingers and not on dashboards where everything should be as informative and visual as possible.
Don't know where this need to modernize GUI is coming from, but I would be as happy as it gets if developers just made a 1:1 remake of old GUI using a new framework to modernize it on the back end, make it snappier, maybe introduce more flexibility to API, but layout design was great. Only thing that could make it perfect is to move graphics to vector based images.
I only hope that this is a draft version of new layout and devs were not really focusing on the look&feel, but on the the back end aspects, and I am just not seeing yet the whole picture of the end result.

 

[KrisBee]

Owners of HP Gen8 Microserver have a boot problem: https://redmine.ixsystems.com/issues/52441
Known for two months.

 

[guermantes]

That should change shortly (we're still in the process of uploading and updating everything).

Perhaps "shortly" is longer than I expected. In any case I am still seeing "This is not a production release, and should only be used for testing."

 

[MatthewSteinhoff]

Tried to go from RC2 to Release. Not having much luck.

Follow-up. Made it to 11.2 Release. Tried again from the legacy interface. Same problem: core dumps and no upgrade. Tried the upgrade from the new interface and it worked.

Cheers,
Matt

 

[ThreeDee]

Follow-up. Made it to 11.2 Release. Tried again from the legacy interface. Same problem: core dumps and no upgrade. Tried the upgrade from the new interface and it worked.

Cheers,
Matt

hmm .. I will try upgrading from new gui and see if that cures my woes .. Thanks!

 

[Spearfoot]

I upgraded my test system from 11.1-U6 to 11.2-RELEASE without any hitches a couple of days ago. So far, so good... I don't use any kind of FreeNAS-based virtual machines, which seems to be the big stumbling block for a lot of people.

Been playing with the new UI, which seems pretty slick.

Being the cautious sort, I'll probably wait until a couple of smooth point releases come out before migrating my two 'production' machines here at home, or any of the servers at work.

Kudos to the developers; they've obviously worked very hard to make this happen.

 

[cods69]

I have not seen any posts from anyone updating from 9.10.2-U6. Any good/bad news stories anyone?
My setup is pretty vanilla, but just interested to see if anyone else is finally taking the plunge out of 9.10.2 into the brave new world.
I'll eventually get around to a test myself, but will take some serious time to back everything up.
It's a prod machine and I need to be cautious, especially with the not-quite-insignificant amount of failures I'm reading in this thread and others.