Unraid OS version 6.12.0-rc5 available

I have been testing RC5 since it released, haven't noticed any issues so far (I did have a few hiccups with RC4, but I believe those were addressed or I haven't noticed them) and it's been rock-solid when combined with ZFS Master (for dataset interface) and Sanoid (for Snapshot scheduling).

 

 

7 hours ago, dickesW said:

Hi unraid Team,

 

I am setting up a new unraid system and want to use ZFS as FS for array discs from the start. Copying the first test files via SMB brings my ZFS memory utizlization from 3% to 74% and stays there. I have 32 GB RAM installed and the array is 6 TB SSD only.

 

Thx dickesW

 

unraid-diagnostics-20230505-1014.zip

 

That'd be the ZFS ARC cache, you can adjust the usage, but as another member said it's probably 74% of the available ZFS (RAM) cache.

 

ZFS has 2 caching tiers. 

- ARC: Uses upto X% of available RAM (I think unRAID has this set as 1/8th of the total system RAM, you can adjust it)

- L2ARC: This can be enabled to act as a secondary (larger, but slower) cache, usually placed on an NVME/SSD drive -- as items fall out of the ARC cache, they can end up in L2ARC if enabled.

- These are considered volatile. Not a big issue with read workloads, but worth noting for write heavy workloads. ZFS groups transactions into a memory buffer and then writes to disk. So you want some form of battery backup, by default I think the write buffer fires every ~5 seconds. Again, not a concern if you're doing low-writes and heavy-reads (like a media server).

 

So as you read data it gets loaded into the ARC cache and stays until it's evicted. If you have L2ARC enabled, it would then move to that. This is why ZFS is really good at read access / heavy read workloads, because after the first I/O hits, it's stays in hot memory.

 

7 hours ago, isvein said:

Im trying to understand the ZFS lingo:

 

-an Vdev is an set of drives in a given raid array (can be compared to a raid array like raid0,1,5,6 etc)

 

-an ZFS Pool is an logical collection of 1 or more Vdevs that looks like a single point of access to the user

 

-Vdevs are independent even if in a ZFS-Pool, so say 1 Mirror Vdev and 1 stripe Vdev in same ZFS-Pool does not make it equal to raid 10

 

-if one drive in any Vdev goes down, you loose access to the entire ZFS pool

 

Do I understand that correct?

 

JorgeB gave a great answer. Once you're past the Pool/Vdev config, you have datasets which appear as just normal folders, but are a collection within the larger pool. You can also nest datasets within datasets if you want to get granular with your snapshots. 

 

For example, on my system I have 4x4TB drives configured into 1 ZFS pool with 2 vdev mirror groups of 2 drives (~7TB of usable space). 

 

- A few side notes, LZ4 compression is really good, and there's been some tests that show it can perform faster (I/O wise) when enabled. 

- If you're worried about redundancy, Vdev mirrors are the way to go. They can be a tad more friendly. See: https://jrs-s.net/2015/02/06/zfs-you-should-use-mirror-vdevs-not-raidz/

- To the end-user, this just looks like a folder (zfspool) of folders.

 

You can get an idea for the flexibility of this in the screenshots below, using the Sanoid plugin to handle scheduling of  snapshots of sub datasets for individual docker/VM's running. Interface is from the ZFS Master plugin. I imagine a lot of these plugin features will eventually make their way into native unRAID at some point, since ZFS already provides this stuff - it's just the interface.

 

 

Edited May 5, 2023 by KyleK29

I hope 6.11 to 6.12 transparently migrates everything like always.

With all these fundamental changes, I wonder if 6.12 is the correct versioning. I can easily be named 6.20.

 

I am still confused on a few things:
1) Functionality of UNRAID WITHOUT using the new ZFS features. We are not loosing anything or getting forced to re-create anything?

2) Functionality of UNRAID WITH new ZFS features AND ability to migrate easily (we need extra disks?).

3) Primary/Secondary storage vs. Array and Cache... I am still confused which is which and if anything really changes aside of the names (AND things really migrating the old to new).

 

1. If you don't use ZFS, existing array and pools continue as before.
2. I haven't used ZFS yet. Might consider converting my 2x500G btrfs pool named 'cache' to ZFS. Everything would have to be copied off so the pool could be recreated as ZFS.
3. Primary is where new files get written. Secondary is where overflow goes. A separate setting controls which direction mover moves (pool <--> array). When you upgrade, it all still works as before since these settings are functionally the same as what we already had. See my earlier posts about this in the rc4 thread

 

https://forums.unraid.net/bug-reports/prereleases/unraid-os-version-6120-rc41-available-r2372/?do=findComment&comment=23380

 

https://forums.unraid.net/bug-reports/prereleases/unraid-os-version-6120-rc41-available-r2372/?do=findComment&comment=23382

 

Hi all, been having this same issue since the 6.12 RC's started, only just now getting around to it. I've seen a few other posts that a similar but none have the same as my logs and none of the fixes seem to work (ie changing to macvlan).

 

Basically after a few hours I cannot access the server at all, not webui, ssh, or ssh directly on the machine. Pressing the power button on the server starts the shutdown process but it just hangs.

 

Only way I could capture anything in the logs was to mirror them to the flash, but here is the lead up to the last 2 times it went offline.

 

Any insight would be appreciated,

May 5 MassiveDynamic.txt May 6 MassiveDynamic.txt

Possibly Pre-Clear plugin issue in RC5

 

Day before yesturday I was trying to setup new server using 6.12.0-RC5. 3 out of my 6 disks showed as failed at the begining of Post-Read state . I read that it can be a sign of memory failure so I ren memtest86. No memory problems. Then I downloaded 6.11.5, I am around 55% Post-Read with no issues. I think that there may be an issue with the Preclear on RC5. I am sorry I didnt think to download Diagnostics file, it is gone by now. 

Edited May 5, 2023 by mwodzynski
Added 1st line

Hello everyone,

my english is not so good, so i use google translator.

I recently updated my Unraid to 6.12.0-RC5. Then I switched the pool to ZFS and then all the disks in the array. The conversion went smoothly with the recommended procedures.

But now I've noticed that the disks in the array no longer go to sleep (spin down), although all files that are used regularly (appdata, system, domain) are in the cache in the pool. On the "Start" overview page I can also observe again and again that write and read accesses take place.

What I also noticed is that the free memory of the network drives on my Windows computers is no longer displayed correctly. And also the output of "df" looks "strange". I have no idea if this is related.

I would be very happy about feedback.

Greetings
Jochen

Ausgabe df.txt brutus-diagnostics-20230506-0912.zip

9 hours ago, docbrown said:

Hi all, been having this same issue since the 6.12 RC's started, only just now getting around to it. I've seen a few other posts that a similar but none have the same as my logs and none of the fixes seem to work (ie changing to macvlan).

 

There's nothing relevant logged, the macvaln issue usually leaves macvlan related call traces in the syslog, were those logs saved after a crash?

8 hours ago, mwodzynski said:

Possibly Pre-Clear plugin issue in RC5

Please report this in the existing preclear support thread instead.

34 minutes ago, JorgeB said:

the macvaln issue usually leaves macvlan related call traces in the syslog, were those logs saved after a crash?

It would appear at some point I switch back to ipvlan I guess in an attempt to try something else. For now I'll leave it and wait for the next crash. For now I've also disabled ipv6 as I noticed that a lot in the logs so I'll see how that holds up before switching back to macvlan

3 hours ago, Jochen Kklaus said:

What I also noticed is that the free memory of the network drives on my Windows computers is no longer displayed correctly.

Some shares do look wrong, but cannot reproduce or see why it's happening in the diags, will require further investigation or mores cases to see what they have in common.

 

As for the spin down there are been some reports before, but nothing conclusive yet, did you do a new config at any point with v6.12?

18 hours ago, KyleK29 said:

 

JorgeB gave a great answer. Once you're past the Pool/Vdev config, you have datasets which appear as just normal folders, but are a collection within the larger pool. You can also nest datasets within datasets if you want to get granular with your snapshots. 

 

For example, on my system I have 4x4TB drives configured into 1 ZFS pool with 2 vdev mirror groups of 2 drives (~7TB of usable space). 

 

- A few side notes, LZ4 compression is really good, and there's been some tests that show it can perform faster (I/O wise) when enabled. 

- If you're worried about redundancy, Vdev mirrors are the way to go. They can be a tad more friendly. See: https://jrs-s.net/2015/02/06/zfs-you-should-use-mirror-vdevs-not-raidz/

- To the end-user, this just looks like a folder (zfspool) of folders.

 

You can get an idea for the flexibility of this in the screenshots below, using the Sanoid plugin to handle scheduling of  snapshots of sub datasets for individual docker/VM's running. Interface is from the ZFS Master plugin. I imagine a lot of these plugin features will eventually make their way into native unRAID at some point, since ZFS already provides this stuff - it's just the interface.

 

 

Now it got complicated 😵‍💫

 

But thanks 😃

upgraded from 6.11.5 to 6.1`2.rc5.

after start array unraid shows me that every disk in arrays is "returned to normal utilization level", what is weird because most of the disks are filled over 80%, notifications is set to 80% and on critical; 90%.

But now every disk is showing green...

rest is running fine 👌

2 hours ago, JorgeB said:

Some shares do look wrong, but cannot reproduce or see why it's happening in the diags, will require further investigation or mores cases to see what they have in common.

 

As for the spin down there are been some reports before, but nothing conclusive yet, did you do a new config at any point with v6.12?

Yes I have because I took two small hard drives out of the array. But I think that was with RC2 or RC3 and before I started changing the disks to ZFS.

Ausgabe df.txt

Edited May 6, 2023 by Jochen Kklaus

18 hours ago, Jochen Kklaus said:

Yes I have because I took two small hard drives out of the array.

There have been some reports that doing a new config, especially with rc2, could cause spin down issues, for some it helped going back to 6.11.5 then upgrading back, others going back to v6.11.5, do a new config and then upgrade back, zfs won't mount with v6.11.5 but you can still try either of those to see if it helps, of course don't format the disks when they come up unmountable.

I have 2 issues:

 

- Docker icons go grey.

- Dashboard regrouping do not stick.

1 hour ago, frodr said:

Docker icons go grey

Screenshot?

23 minutes ago, trurl said:

Screenshot?

 

 Starting a docker, them the coloured icons are back for a while. I can't see what's causing the grey.

 

 

 

 

Dashboard: I can re-arrange and the lock it. But go away and the back again and the changes are gone.

helmaxx-diagnostics-20230507-1749.zip

Edited May 7, 2023 by frodr

You system / network is unable to download the icons after the first installation.  (After it downloads once, they are cached)  Alternatively, your docker image is very corrupt.

 

For the dashboard, does this happen if you close the browser down?  Are you automatically clearing cookies when closing the browser?

25 minutes ago, Squid said:

You system / network is unable to download the icons after the first installation.  (After it downloads once, they are cached)  Alternatively, your docker image is very corrupt.

 

For the dashboard, does this happen if you close the browser down?  Are you automatically clearing cookies when closing the browser?

 

I don't think docker image is corrupt as it happens on both server. 

 

For dashboard it happens when going dashboard - main - dashboard. Not closing the browser and clearing cache unless specifically needs to. 

 

I will clear cache and see what happens.

 

12 minutes ago, frodr said:

 

I don't think docker image is corrupt as it happens on both server. 

 

For dashboard it happens when going dashboard - main - dashboard. Not closing the browser and clearing cache unless specifically needs to. 

 

I will clear cache and see what happens.

 

 

Cleared cache. Rearranged dashboard lost after dashboard - main - dashboard. Ot just updating browser page. 

Edited May 7, 2023 by frodr

how do I get the diagnostic when my system crash? after updating I have been getting kernel panics. I tried rebooting but the log was gone. this is how the panic looks 

10 minutes ago, theangelofspace15 said:

how do I get the diagnostic when my system crash? after updating I have been getting kernel panics. I tried rebooting but the log was gone. this is how the panic looks 

Use the Syslog Server:

     https://forums.unraid.net/topic/46802-faq-for-unraid-v6/page/2/?tab=comments#comment-781601

 

3 hours ago, Frank1940 said:

Use the Syslog Server:

     https://forums.unraid.net/topic/46802-faq-for-unraid-v6/page/2/?tab=comments#comment-781601

 

Thanks, Now I have it setup with a share called `logs` Let see if if happens again. This is the diagnostic after the server rebooted.

unraid-diagnostics-20230507-1906.zip

45 minutes ago, theangelofspace15 said:

Now I have it setup with a share called `logs` Let see if if happens again

That won't work.  It's a common misconception of the syslog server.  The share option is for where to send logs that are incoming (eg from another server).  You want the mirror to flash