Kernel Panic on Mac Mini

Waiting is the only solution, until Apple releases a fix, unfortunately. Hopefully they made progress over the holidays, we won't hear until next friday. When there is a possible solution, it will be in a MacOS beta release hopefully.

Its incredibly frustrating that this obvious and devastating bug has lasted way over a year, without resolution. We have no way to work around it, except setting the 64k stripe unit size.

@softraid-support 

if this bug is known for over a Year wat was default setting in softraid 7 still at 16k and not 64k, at least that  would have bought time to wait for Apple update? 

@patbarollo 

There are lots of reasons for this.

We hear month by month, "its almost fixed". We have production (manufacturing) code that would take a long time to change, and the large majority of users never experience this.

and 16k is more efficient for most users, so we want to keep that the default option.

what's worse is ventura seemed to have fixed this (or most all of it) during the beta period, then some of these cases came back with the release. Whether a new bug was introduced, or what, we do not know.

In the meantime i have contacted Apple support in the hope to apply pressure. No dice. The support collaborator that contacted me made clear that a far as he thought Apple would not adapt their OS and it would be up to Softraid to find a workaround.

I was to contact Softraid support and Softraid needed to handle this with the help of Apple.

Can you please give us an idea how the contact between Softraid and Apple are progressing concerning this issue ?

This is not an isolated issue, a lot of users are experiencing this problem.This makes the OWC enclosure and softraid an unusable solution and bad investment to the extent that I wonder if I should not request a refund from OWC in the amount of the damage incurred ?

@patbarollo 

the person you spoke to would not know anything. We are working with Apple engineering directly. With the holidays, nothing has happened recently, however.

Did you convert your RAID 5 volume to 64k stripe unit size to work around this?

@softraid-support 

Wait, what is this bug?  My raid is configured with 16k and from my other thread you have seen I am having all sorts of issues, is this related?

@chrispeden 

Here is how to tell:

If you are running an ARM computer and have RAID 5 with 16k Stripe unit size, maybe.

If your system panics, then "report to Apple", click "show details", and either save the support file (as Make Plain Text in Text Edit, so you can attach it, or look at the top lines, and see if it shows "DART". Then you are being hit by this DART IO controller issue.

Like i said before, since you did not mention this recommendation in the installation notes and softraid uses the default value of 16K stripe i would need to move 80 TB of data to an older NAS what would take +6 days , delete and recreate the volumes with 64K stripe and then  move the data again from the nas to the thunderbay (again 6 days). So if a start today and nothing goes wrong i would be able to tell you if this worked by the end of january. I would really like a workaround or solution that permits me to avoid this. And i read from several people that they experience the problem even with the 64K stripe ? So am i certain this is going to work ?

@softraid-support 

2 out of the 3 kernel panics I posted are indeed that.  What is the con's if any to recreate my raid with 64k?  I need stability more than anything and this is killing my ability to get my work done.

@chrispeden 

THe only con is really small files will be slower to load, not much you would ever notice.

@patbarollo 

There are actually very few (but non zero) that have seen this with 64k.

@patbarollo @softraid-support

I took the plunge.  I am verify 100% my problems have went away.  (not sure about random spin down / sleep yet) but I have had ZERO kernel panics since recreating my RAID 5 with 64k stripe size.  The whole computer seems more stable in fact.  I am not sure how or why a stripe size would have any bearing on this but it indeed does.

@chrispeden

My theory is that SoftRAID is multi-threaded, and the small IO's in 16k, along with a bug in the DART processor when trying to manage many small IO's is what is causing this. I am not an engineer, however. 64k should not be a fix, but it is.

thanks

Any news from Apple support ?