Just curious, how many of you have
switched to QuTS h6.0.0 release candidate
and if you have any regrets
Iāve been using QuTS hero 6 since its launch and itās ok for me (I donāt have any problems) 
Having previous experience, Iām not going to install a beta firmware. Besides, Hero 6 seems to be missing quite a few things from Hero 5.
What is missing?
The following applications will no longer be supported in QuTS hero h6.0.0: CAYIN CMS-WS Lite and CAYIN MediaSign Player (replaced by CAYIN Media Viewer), IDrive, JRE, Mustang Card Manager, Mustang Card User Driver, Python, Python3, QButton, Qmiix Agent, QVR Elite (replaced by QVR Surveillance)
All details here: QuTS hero h6.0.0.3459 build 20260409 Release Candidate | Release Notes | QNAP
3 of those (CAYIN CMS-WS Lite, CAYIN MediaSign Player, QVR Elite) have been replaced with 2 QPKGs (CAYIN Media Viewer, QVR Surveillance) and are still supported in QuTS 6.
There are Python QPKGs available in the myQNAP repository.
The Mustang card management apps are maintained by QNAP. If theyāre not going to support the cards in QuTS, then the QPKGs wonāt be needed either.
QButton? No loss there.
Qmiix Agent? Some loss, but being home automation, I imagine thereās something else out there that already performs the same function.
IDrive? Still supported in QuTS 6 according to the release notes.
That leaves JRE. If you need it, ask the myQNAP folks if they can create a new package for it.
If iDrive is going to be supported in QuTS 6 thatās a big deal for me.
Personally i would wait until the 2nd or 3rd official release before considering this update, software seems to be released in an unfinished state nowadays !
I have decided to no longer wait because I need Qtiering (Qtier) on ZFS. If you create a Qtier with QTS - which is what I did during QuTS hero 6 beta stage - you will have to delete the entire QTS storage pool (EXT4/LVM) and recreate it under QuTS hero (ZFS). In parallel Iāve been running a Qtier on ZFS and QuTS hero already since the QuTS hero 6 betas and I have not had problems. The only limitation with Qtier on/under ZFS is: Sync between hot data and cold data, i.e. SSD RAID and HDD RAID only happens once every day and not dynamically like is set by default on Qtier on QTS.
Since data integrity is handled by ZFSā (super) robust architecture, I donāt see why Qtier for ZFS should cause problems aside from potential inferior performance in case the hot/cold swapping routines are not yet as mature as on QTS where Qtier now is available for 10 years. When copying really huge single files, e.g. 200GB - 400GB, you can expect about 20 - 25% less performance (over 10G and SMB) when compared with a RAID 10 or RAID 6 (granted, it is not filled for more than 75%). But as soon as files are smaller or like a lot smaller, the benefits of Qtier start to pan-out unless the SSD RAID is not completely filled with hot data already before the daily shifting to the cold HDD RAID occurs.
In the near future I expect that Qtier on ZFS will shift hot SSD data to cold HDD data dynamically too. In the meantime I can live with the once a daily-shift. In case you have a QNAP that really needs high performance for the entire day for hot data, the SSD RAID of the Qtier should be bigger. E.g. if your SSD RAID of the Qtier on ZFS is configured as RAID 1, you can perform a disk by disk, actually SSD by SSD replacement with bigger SSDs. This is possible under Qtier too because Qtier is nothing else then a ZFS Storage Pool with two or three RAIDs (NVMe SSD, S-ATA SSD, HDD). I have also seen that you can expand a RAIDZ1 (RAID 5) or RAIDZ2 (RAID 6) under āActionā with incremental HDDs or SSDs. So, while I have not yet tested this particular scenario because I have no incremental HDDs in this QNAP, I expect this option to work with a Qtier on ZFS too because the option āExpandā under āActionā is not greyed out. Thus, like starting with OpenZFS 2.3, you can expand a RAIDZ1 and RAIDZ2 with incremental SSDs or HDDs. It does not work if you create a RAID1 only, initially for the hot RAID of the Qtier. In that case you must do a one by one SSD replacement with bigger SSDs. Under āActionā I see the option to create/add incremental RAIDs to the Storage Pool but I donāt know if Qtier can properly handle more than three RAIDs (hot - warm - cold). Still, adding more HDDs later to the cold RAID or adding more SSDs to the hot or warm RAID later if those are configured as RAIDZ1 (RAID 5) or RAIDZ2 (RAID 6), that seems to me to be no problem because this is handled by OpenZFS 2.3 or later and not by the Qtier routines.
Thus, you are not stuck with a Qtier to the initial size forever. You can later expand a Qtier too, namely the cold HDD RAID. BTW: You might want to be careful: I have seen that currently QuTS hero 6 does not prevent you from creating a hot (NVMe U.2 or M.2) or warm (S-ATA SSD) RAID as a Stripe (RAID 0). Donāt do this! If your SSD Stripe fails, the entire ZFS Storage Pool is lost. Thus, your hot and/or warm RAIDs of the Qtier must be configured as RAID 1.
You might reasonably ask: Why putting such an emphasis on Qtier on ZFS?
Because even an all Flash S-ATA storage-pool with some 20TB is not affordable anymore. The 3.84TB Kingston DC600M is at ā ā¬/$1,500.- per piece. Thatās 9,000.-- for less than 20TB net capacity and only RAID 5 and only the SSDs, NAS not included in the calculation.
Thus, in this most severe storage crisis, Qtier becomes like a life-saver because only hot data that needs SSDs is being stored on SSDs and all cold data is being moved to HDDs. Sure, Iām aware that hot & cold data not always is to distinguish straight forward but Qtier already helps a lot. It would be a nightmare to have to shift data manually between the SSDs and HDDs. And if you were to create a standalone SSD Storage Pool and a HDD Storage Pool, you would even have to copy data between the SSDs and HDDs, not just moving!
whats qnapās end goal with the three version of quts hero? Are 5.2/5.3 going away and only 6.x will remain? Iām trying to understand why there are three versions of the OS out there.
5.3 is really only meant for those who need High Availability. I think itās a bit of a ābridgeā release between 5.2 and 6. I expect that once 6 is released, you will begin to see just maintenance releases (security updates, etc) for 5.x just like you would with any other āoldā releases.
Iām hesitant to move my production TS-1655 over to 6.x because Iāve had ongoing SMB on macOS issues with QuTS hero 5.x.
My (intermittent) symptoms include macOS Finder error -36, occasional āFile Not Foundā behavior on NAS paths, temporary/ghost files such as .sb-xxxx files being created, and SMB instability during heavier copy/hash/metadata workloads over SMB.
I see h6.0 release notes mentions improvements for Kernel Mode SMB Daemon with Encryption, but nothing that says SMB / macOS stability issues being addressedā¦
Has anyone experienced these mac / SMB issues on 5.x? And any improvements or changes with SMB reliability on h6.0?
What version Mac OS are you running? I have had no issues.