TrueNAS CORE 12.0-U6 Released

[wsoteros]

TrueNAS CORE 12.0-U6 has been released!

Read the changelog: https://www.truenas.com/docs/releasenotes/core/12.0u6/

Download: https://www.truenas.com/download-truenas-core/


Let us know what you think below!

 

[awasb]

Installed fine on my test system. No issues so long.

Just somer minor glitches in reporting / theme preferences etc.

(jira tickets are filed)

https://jira.ixsystems.com/browse/NAS-112690

https://jira.ixsystems.com/browse/NAS-112705

 

[elvisimprsntr]

Missing CPU temp data on reporting page, apparently known issue

https://jira.ixsystems.com/browse/NAS-112690

 

[Phil1295]

Whenever I login to GUI (I am using 2FA with Google Auth), I am prompted with a "Get Started" Welcome annoying message. There is no option to set it to never popup again !
Is it a bug or a step to nag free users ?

 

[jlpellet]

Regarding CPU reporting page, 3 of my 5 systems have the issue. Other 2 CPU report page is correct. Regarding annoying splash screen, I see it only for 1st loging but am NOT using 2FA.

 

[Phil1295]

Whenever I login to GUI (I am using 2FA with Google Auth), I am prompted with a "Get Started" Welcome annoying message. There is no option to set it to never popup again !
Is it a bug or a step to nag free users ?

In addition, the columns layout is no longer saved and the popup to "Show hidden users/groups" is displayed each time I access the users/groups pages after a logout / reauth

It was not the case on last U5.1

 

[HoneyBadger]

When ZFS dedup is enabled on a pool, use SHA512 as the checksum algorithm

Question; I'm pretty sure changing the checksum algorithm will cause the existing deduplication tables to return "no match" against the previous one.

Eg: If you were previously using SHA-256 and create a new dataset with SHA-512, it will have zero deduplication against the SHA-256 dataset as the hashes will never match.

This seems like a bit of a dangerous thing to change ... is there a way to scan the active deduplication table on a pool to query the algorithm in use?

 

[vikonen]

Backup system upgraded succesfully. Everything seems to work fine, except the Reporting tab shows an empty graph for CPU Temperature.
Unfortunately I did not check whether it was empty prior to the upgrade.
On the Dashboard page it shows the temp for the hottest core and the bar for CPU temp correctly.

 

[vikonen]

I checked that the CPU temp graph works fine on the main system prior to upgrade from 12.0.U5.1 to U6. After upgrade it no longer works. I suppose a bug ticket is in order.

 

[Samuel Tai]

I checked that the CPU temp graph works fine on the main system prior to upgrade from 12.0.U5.1 to U6. After upgrade it no longer works. I suppose a bug ticket is in order.

No need for a new bug ticket. One's already in work. https://jira.ixsystems.com/browse/NAS-112690.

 

[Ericloewe]

Question; I'm pretty sure changing the checksum algorithm will cause the existing deduplication tables to return "no match" against the previous one.

Eg: If you were previously using SHA-256 and create a new dataset with SHA-512, it will have zero deduplication against the SHA-256 dataset as the hashes will never match.

This seems like a bit of a dangerous thing to change ... is there a way to scan the active deduplication table on a pool to query the algorithm in use?

I expect that's a new default at activation time, not a retroactive change.

 

[Jailer]

I expect that's a new default at activation time, not a retroactive change.

It is, it's in the release notes.

[NAS-110523] - When ZFS dedup is enabled on a pool, use SHA512 as the checksum algorithm

 

[HoneyBadger]

I expect that's a new default at activation time, not a retroactive change.

Deduplication tables are pool-wide though, so in the example of existing datasets/zvols with SHA256 in use, creating an additional dataset/zvol using SHA512 will have zero overlap.

 

[Etorix]

So in the case of a pool where dedup is already enabled no change shall be made to the pool-wide setting and additional deduped datasets will keep using the checksum algorithm that is already in force.
If that's not the case: "iXSystems, we've had a problem…"

 

[Etorix]

No need for a new bug ticket. One's already in work. https://jira.ixsystems.com/browse/NAS-112690.

Upvoted!
Even if it's just a cosmetic issue, that an update introduces a regression on a previous fix does not bode well for the validation process and quality control.

 

[Samuel Tai]

Upvoted!
Even if it's just a cosmetic issue, that an update introduces a regression on a previous fix does not bode well for the validation process and quality control.

From what I can see in /usr/local/lib/python3.9/site-packages/middlewared/plugins/reporting/plugins_freebsd.py (which incidentally is identical between U5.1 and U6), this is a latent bug in U5.1, or possibly earlier, which unfortunately triggered in U6. The relevant lines are line 22, which gets the number of CPU cores, and returns a 1-indexed value, and lines 29 and 45, which ingest those number of cores to use as a loop index to pull the RRD CPU temperature files in /var/db/collectd/rrd/localhost/cputemp-{n}/temperature.rrd. However, as loop indexes, the loops expect a 0-indexed value, so the number of cores has to be decremented by 1; otherwise, there will be a non-existent path at the last loop, which results in no data being displayed. Alternatively, the loops can be reindexed to start from 1.

 

[Ericloewe]

Deduplication tables are pool-wide though, so in the example of existing datasets/zvols with SHA256 in use, creating an additional dataset/zvol using SHA512 will have zero overlap.

Yeah, that's an edge case that may catch someone by surprise, but it's like a papercut on somebody who's been in a car crash.

 

[Phil1295]

Whenever I login to GUI (I am using 2FA with Google Auth), I am prompted with a "Get Started" Welcome annoying message. There is no option to set it to never popup again !

In addition, the columns layout is no longer saved and the popup to "Show hidden users/groups" is displayed each time I access the users/groups pages after a logout / reauth

It was not the case on last U5.1

Anyone else having this issue with 2FA
I did not try to disable 2FA, but it is really annoying, should I file a bug or it is a known issue ?

 

[Alecmascot]

it is a known issue ?

If it is a known issue there will be an open bug for you to search for .

 

[Phil1295]

I opened a ticket: https://jira.ixsystems.com/browse/NAS-112744
Hope someone can point to a fix doable by shell to fix the issue. Seems like the GUI parts are not saved