Cant start up MacStudio with Thunderbay 4 attached

Hi, personally, I still have the same problem, and I can't use the OWC ThunderBay 4 mini (for two months now) because my Mac Studio crashes immediately.

My configuration: SoftRaid 7, macOS Ventura 13.0, Mac Studio M1 Ultra, OWC ThunderBay 4 mini with 4x NAS SSD, RAID 5 with 64k stripe unit size

Is there already a solution please? :) Thanks

@jirka 

Have you ever been able to use it? (No data on the disks?)

@softraid-support Hi, thanks for the reply. Yes, if I didn't use RAID 5, or if I connect the drive about 10 minutes after starting the computer. Then it works for me. If I plug it right after startup or it's pre-plugged, the Mac crashes. Also, the method where one drive is ejected and I plug it in later works. But none of this is a solution. I thought MacOS Ventura and SoftRaid 7 would fix the problem :)

@jirka 

Ventura beta's did fix this. then the release brought some back. I don't understand if it is the same issue, partially resolved, if the "fix" was backed out, or if it was modified in some way that made it less effective.

In early beta testing in Ventura, all the testers reported it was fixed. Very frustrating!

@softraid-support Hi, just upgraded to MacOS 13.0.1 and the problem remains. OWC Thunderbay is still unusable, 3 months already :(

@jirka 

Please do not post the same topic in multiple threads. You won't get a response any faster.

Are you willing to do a clean install, on a second system in your internal drive? It won't affect your other system.

Run Disk Utility. Select your existing ovlume, then click + volume.
shut down, restart in recovery mode (hold the power button)
do NOT migrate data.

"reinstall MacOS." Select the new volume.
when done, you have to do a trick:
set start to the old volume, restart
Set Startup to the new volume, restart
Now you can set System Security as before from Recovery mode.

install SoftRAID, etc.

Does this let your volume mount?

@softraid-support I wrote on other topics not for support, but to let the authors know that there are other users with the same problem :) I have already done what you write. I bought a new Mac Studio from Apple, bought a new OWC thunderbay 4 mini from OWC and plugged in brand new SSD drives. RAID 5 was the first thing I dealt with after turning on my new Mac. And kernel panic was an immediate problem. It must be a bug between SoftRaid and Apple Silicon :/

@jirka 

Lets avoid multi-posting, though. It makes managing the forums much harder.

What we found in testing was the panic was triggered in the System. Did you "migrate" data from a prior system?
One of the tests I did was obtain a users crashing system. Any RAID 5 volume crashed. Creating a clean install, did not crash, so the trigger has to be inside the System, that is crashing the DART IO controller. Other users confirmed this, they could wipe their drives, no change, but a clean install, and it goes away. Others have reporting this suddenly appearing on a previously working system.

We are working as hard as we can to get a fast resolution.

@softraid-support Thanks for the additional information 👍

Just a quick note: after updating SoftRaid to version 7.0.1, the problem remains. In the release notes there is "Fixes a kernel panic which occurred when using encrypted APFS volumes.", if I change the storage from APFS to encrypted APFS... will that solve my kernel panic problem? Thank you!

@jirka 

Yes. We did not announce support for encryption yet, as we are working on adding this feature to the user interface. but encrypted volumes should work fine now.

@softraid-support thank you for working on a solution, I will wait for the next update and try switch to encrypted APFS Raid5 volume.

@jirka 

Good!

The only way to do it currently, is erase the volume with Disk utility.

This just happened to me today. After attempting to copy several terabytes of data my machine crashed repetitively, had to turn off the enclosure. Does even creating a 64k stripe set resolve this?  My system is barely usable if I leave it on. 

@jawong

Yes, changing to 64k fixes this over 95% of the time, I hate to claim 100%, as there have been a couple outliers.