Register for the iXsystems Community to get an ad-free experience and exclusive discounts in our eBay Store.

TrueNAS 12.0-U2 now available!

Western Digital Drives - The Preferred Drives of FreeNAS and TrueNAS CORE

o0o

Junior Member
Joined
Feb 10, 2021
Messages
20
You can look at the contents of /var/db/system/cores by running the following command from shell:
ls -l /var/db/system/cores

Some cores may be quite old from previous releases. These cores can be removed by running the command
rm /var/db/system/cores/<file>

If you see an smbd.core in there it's nothing to worry about and can be safely removed. There will be quite a few people with one because of a bug present in U1 and fixed in U2.
rm /var/db/system/cores/smbd.core

If you see cores that (1) aren't ancient and (2) aren't smbd.core, syslog-ng.core, or winbindd.core, you can tar them up and attach to a bug ticket
tar -cvzf /tmp/cores.tgz /var/db/system/cores (then use scp/sftp to copy it off). Since a core file may contain sensitive info, be sure to mark the upload as private.
Great, that worked and was, like you described it. THANKS! :smile:
 

greentin

Newbie
Joined
Jan 24, 2021
Messages
1
Hi,

First of all, thanks for all the incredible work on this release and Truenas.

I have another .core file that the one you mentioned. It is libvirtd.core.
I'll open up a bug ticket.

Kind regards
 

mgittelman

Senior Member
Joined
Dec 8, 2017
Messages
404
You can look at the contents of /var/db/system/cores by running the following command from shell:
ls -l /var/db/system/cores

Some cores may be quite old from previous releases. These cores can be removed by running the command
rm /var/db/system/cores/<file>

If you see an smbd.core in there it's nothing to worry about and can be safely removed. There will be quite a few people with one because of a bug present in U1 and fixed in U2.
rm /var/db/system/cores/smbd.core

If you see cores that (1) aren't ancient and (2) aren't smbd.core, syslog-ng.core, or winbindd.core, you can tar them up and attach to a bug ticket
tar -cvzf /tmp/cores.tgz /var/db/system/cores (then use scp/sftp to copy it off). Since a core file may contain sensitive info, be sure to mark the upload as private.
I got the same error. I have sshd.core file present. Is it safe to delete?
 

j_r0dd

Member
Joined
Jan 26, 2015
Messages
109
I do get some graphical issues in the webui dashboard for the cpu cores. Been getting ever since installing 12
 

Attachments

Samuel Tai

Never underestimate your own stupidity
Moderator
Joined
Apr 24, 2020
Messages
2,695
I do get some graphical issues in the webui dashboard for the cpu cores. Been getting ever since installing 12
Standard procedure for GUI issues after upgrades is to clear your browser cache and history.
 

j_r0dd

Member
Joined
Jan 26, 2015
Messages
109
Standard procedure for GUI issues after upgrades is to clear your browser cache and history.
I did clear cache and history. I probably should have tested on other browsers before posting. Looks to be an issue specific with Safari. Firefox looks fine.
 

zamana

Member
Joined
Jun 4, 2017
Messages
115
Hi!

I waited until the 12.U2 to upgrade from 11.3-U5, and it seems that everything went fine.

The only issue until now is that my plex jail (not plugin) is not starting:

Code:
root@freenas[~]# iocage start plex
No default gateway found for ipv6.
* Starting plex
plex devfs_ruleset 10 does not exist! - Not starting jail


Removing the entry from the config.json fixed the issue.

Code:
root@freenas[/mnt/system/iocage/jails/plex]# grep devfs config.json
    "allow_mount_devfs": 0,
    "devfs_ruleset": "10",
    "min_dyn_devfs_ruleset": "1000",
    "mount_devfs": 1,


Thanks.
 

Samuel Tai

Never underestimate your own stupidity
Moderator
Joined
Apr 24, 2020
Messages
2,695
Don’t edit the JSON. Use iocage set devfs_ruleset=4 plex from the shell to modify the property for the current jail.
 

freph

Newbie
Joined
Feb 12, 2021
Messages
2
Upgrade was done from U1 where it was stable for 26 days with 0 issues. For some reason, I get unscheduled reboots on this build after 1hr 10m~1hr 15m of uptime. Entire ESXi host has to be restarted otherwise attempts to reboot hang at mps0 failing with "Doorbell handshake failed".

Hardware:
R720xd
12x1TB drives, 2xRAID-Z2 VDEVs
192GB RAM (tests clean on memtest, no errors from iDRAC)
ESXi-6.7.0-20201104001-standard (Dell, latest)
VM is allocated 8 CPUs and 96GB RAM, RAM is reserved due to a LSI2008 (IT mode) being passed through

Oddly enough, if I drop the VM down to 64GB RAM it will stay up (idea from https://www.truenas.com/community/threads/doorbell-handshake-failed.74800/) . 96GB is obviously overkill for a 10.9T pool, but I don't really see any reason why it would cause the system to crash into an unrecoverable state on U2 and not U1.

Reverted to U1 w/ 96GB RAM and it's back to its usual stable self, but will be staying away from U2 until an issue can be identified. More than happy to provide logs/debugs/etc. or open a bug if needed.
 

Newfoundland.Republic

Neophyte Sage
Joined
Jul 2, 2019
Messages
637
@freph - Check the forums and jira for similar reports. If you don't see anything similar, report the issue to Report a Bug. You will have to register.

Things can't get fixed if iX doesn't know about it!
 

Ruff.Hi

Member
Joined
Apr 21, 2015
Messages
244
Just upgraded my dev server ... it is stalled on the reboot

The last line on the screen is ...

ioapic0 <version 2.0> irqs 0-23 on motherboard

Google implies that it is the NEXT item in the boot sequence that is causing the issue. I will boot from the prior version and try to see what the next step is. Is the boot screen blasts written to a file somewhere?

Edit: My prod and dev servers share keyboard, mouse, screen and I use a KVM switch to jump between the two. The dev server had the ball but it appeared to not be able to 'see' the keyboard on reboot. They keyboard was not responding to numlock toggles.

I cycled the KVM, dev server rebooted (!?) and booted up nicely.
 
Last edited:

Palladio

Neophyte
Joined
Feb 9, 2014
Messages
6
I upgraded with this yesterday with no issues. Thanks!
 

anodos

Sambassador
iXsystems
Joined
Mar 6, 2014
Messages
7,200
I don't know why, but so far each time I have upgraded 12 (from 11 -> 12, then 12 -> 12-U1, and now from 12-U1 -> 12-U2), my Samba users passwords are seemingly reset. That is, in order to get authenticated Samba shares to work after an upgrade, I have to go into Accounts -> Users and set their password again or they are unable to login.
That might be an indication of failure to properly initialize the system dataset (might be overly large amount / size of core files). If system dataset initialization fails, then the setup of samba's passdb database will fail and users will be unable to authenticate. "Resetting passwords" will generate entries in a fresh passdb.tdb file, but this may reside on tmpfs (causing issue to repeat on reboot). /var/log/middlewared.log should be fairly clear if this is happening. If this is the case, then you can run the command midclt call smb.configure rather than resetting user passwords (this will be a backgrounded job in visible in the task manager).
 

Brezlord

Member
Joined
Jan 7, 2017
Messages
177
Performance has improved greatly across the board with 12.0-U2 release. Previously FC was the only protocol that was not showing performance degradation from 11.3. iSCSI is now preforming much better. Bellow tests are done in a windows VM on a pool with 16 running VMs.

12.0-U1.1 iSCSI
12.0-U1.1.JPG


12.0-U2 iSCSI
iscsi_slog_1.png
 

mute

Junior Member
Joined
Dec 8, 2013
Messages
19
That might be an indication of failure to properly initialize the system dataset (might be overly large amount / size of core files). If system dataset initialization fails, then the setup of samba's passdb database will fail and users will be unable to authenticate. "Resetting passwords" will generate entries in a fresh passdb.tdb file, but this may reside on tmpfs (causing issue to repeat on reboot). /var/log/middlewared.log should be fairly clear if this is happening. If this is the case, then you can run the command midclt call smb.configure rather than resetting user passwords (this will be a backgrounded job in visible in the task manager).
I checked and I don't have any cores laying around and while I do have a fair amount of exceptions in my middlewared.log, I don't see anything obviously samba related. I'll continue poking around though to see if I can't get to the bottom of it. Knowing that I can just run midclt call smb.configure should it crop up again is a lot more convenient than having to reset user passwords so thank you for that!
 

Ruff.Hi

Member
Joined
Apr 21, 2015
Messages
244
Just upgraded my dev server ... it is stalled on the reboot

The last line on the screen is ...

ioapic0 <version 2.0> irqs 0-23 on motherboard

Google implies that it is the NEXT item in the boot sequence that is causing the issue. I will boot from the prior version and try to see what the next step is. Is the boot screen blasts written to a file somewhere?

Edit: My prod and dev servers share keyboard, mouse, screen and I use a KVM switch to jump between the two. The dev server had the ball but it appeared to not be able to 'see' the keyboard on reboot. They keyboard was not responding to numlock toggles.

I cycled the KVM, dev server rebooted (!?) and booted up nicely.
Upgraded my PROD install to U2 without a hitch.
 

kikos

Newbie
Joined
Feb 14, 2021
Messages
2
I'm about to install my first TrueNAS box.
I just flashed .iso to 2GB USB-stick with rufus and try to boot (HP DL160G6). Loader reports /boot/lua/loader.lua not found and prompt me with OK. I issued 'boot' command and faced mountroot problem. Now I issued 'cd9660:/dev/da2' at prompt and installation scripts started.

Is it Ok? I'm not sure.
 
Top