SoftRAID Volume Locked error

@steve223

Sorry I have to repeat myself. There is a kernel bug in macOS that affects perhaps .1% of users, and SoftRAID locks the volumes to prevent data corruption.

One day Apple will fix this. It took us 2 + years to get Apple to fix one of the Thunderbolt issues, we have not convinced the right team to apply the engineering resources to fix this disk renumbering issue. Its very frustrating.

It is not a "compatibility" issue. It is a Thunderbolt bug or perhaps macOS kernel bug.

You have one of those configurations that is getting this frequently. I cannot explain why. It could just be interference. Its not common, you are seeing 75% of the cases, as most afflicted users post to the forum.

Every OWC Thunderbay new product goes through the intel/Apple Thunderbay test labs to get certified.

We are in a world where devices are communicating at speeds of billions of cycles per second.

I wish there were a simple fix for these.

Sometimes we try a swap of enclosures, cables are easy to swap, (or can be from a Monitor, USB hub, etc.), but how do you tell a user it is his computer?

as I said I feel for you guy's but it does not change that I wasted 3k on a product that does not work. And by the way it did not work on 2 out of 3 devices 2020 MacPro and as well as 2020 iMac 5K id does work on my 2015 iMac though.

returning items to the US is unfortunately not a good option as it would cost me several hundred $ in postage.

The Promise Pegasus now runs perfectly without any issues on the same machines.

@softraid-support I have a new iMac, and i bought the OWC enclosure at the same time, and added 4 drives to it. They ran about 3000 hours. All of the sudden, within the last 48 hours this issue happens. If this is a bug that is lingering for several years, how do you explain that it happens now? I haven't changed my configuration since i got the enclosure.

@mygiguser

In your case, try resetting NVRAM (command option pr) if it happened starting with a driver update or macOS update. I don't know why, but there is a correlation at times with updates and this happening.

OTherwise, try your cables, remove/insert them a couple times and make sure both ends are tightly connected. Let me know.

@steve223

Good to hear, though.

Question, you are on 240 volts? I wonder if power could have any impact on any of this. These days, I would expect transformers to be extremely solid and reliable one voltage to another , but this makes me wonder.

What country are you in?

Posted by: @softraid-support

@steve223

 

Sorry I have to repeat myself. There is a kernel bug in macOS that affects perhaps .1% of users

 

I'm truly curious how this only affect .1%...how does an MACOS bug only affect some people.  What makes my device different from somebody else's?  There isn't that much variability in Apple products which is supposed to be key to their business model.

And if the affected rate is really that low, then why not refund the customers for their units, not like .1% of your sales is gonna set you back, and the customer service will likely win you some customers.  Sounds like win-win to me.  

@roscoe

The problem is we are dealing with very high frequency signals, so it is likely interference. And the cause is not necessarily the enclosure, but cables, computer, and even macOS itself.

We have replaced enclosures on occasion, and sometimes it works, sometimes it does not, the cause was the computer itself. Every connect point has two thunderbolt chips, and these can clash/crash.

Having a hub can cause this problem. So can having a second monitor.

Upgrading macOS can trigger these events. (and sometimes, resetting NVRAM for some reason can alleviate it)

Also, in macOS, USB is not problem free. This particular problem is not common on USB, except when the user has a hub.

This is a very frustrating issue. What SoftRAID is doing by locking the volume is protecting against data loss from this bug. It is the messenger, not the actual problem.

How does electrical interference relate to the drive renumbering issue?

@roscoe

When you connect a disk, it is assigned a "permanent" BSD disk identifier, like disk5. that can never change, according to the rules of macOS, until the disk is physically removed. then it is assigned a new disk identifier when reconnected.

Two things happen in mac OS that are serious bugs:

the disks eject from the bus because the thunderbolt chips crash

The disks get new bsd numbers, generally waking from sleep.

Both of these are likely to interference. For instance, you can place an iPhone on top of the left port or on teh cable near the power supply and if a call comes in, the disks on that enclosure are likely to eject. That is caused by interference.

This later issue is most likely a crash, or reset from interference, when the computer is sleeping. The disks are assigned new numbers and do not "inform" the IO registry, so data subsequently written to that disk can go to the wrong disk. Heat can also cause the cables to crash, which is a frequent problem with fiber optic cables.

Consider a short Thunderbolt cable, and you can have a megabyte of data on that cable at any "instant".

@softraid-support so apple pretty much killed Software Raid and I say it again you keep selling a product that any Apple devices can have serious problems with without letting your customers know. It cost me several thousand $.

I know it's not your fault but be transparent and tell potential customers about the issue they may be facing!

Honestly, if I'd been told that the problem only affect 0.1% of the units I probably would have rolled the dice and bought them anyway, but given that you know the problem exists (and what you describe sounds like a Thunderbay issue since not all are affected by a common OS), I would expect you to stand behind the units.  You already sent me a replacement and that did not fix the issue, so I'm zero for 3.  If the problem were really 0.1%, then I'm waaaaay out there on the bell curve...too far for me NOT believe this is a Thunderbay issue.  

Note that I'm a professional test engineer who understands quite well Electromagnetic Interference/compatibility [EMI/C] as well as the probabilities of unit failure.  This is on MacSales/OWC to make good with its customers.

@softraid-support given that hardly any Mac has sufficient internal HD space anymore external HD are imperative and need to be working always and without issues if you work of the machine.

on my media centre it's not such a drama but still took me 10 minutes of troubleshooting getting my OWC working again after the latest OS update, just not acceptable for a vital HD in my book.

@roscoe

I do not know your hardware history, but surely you might agree if you have had 3 different enclosures, that you considered the problem is with the mac itself?

Apple knows there is a problem with Thunderbolt. It happens with all hardware, but the more hardware, the more frequently disks eject from the bus.

Lets assume the hardware enclosure is OK, at least for now. (it is a replacement according to your comment)

Have you tried a clean install of Big Sur as a test?Its worth trying.

Believe me it would be frustrating,  but yes it certainly could be the computer. Outside of replacing it, or having a spare, this is difficult to test.

@steve223

I totally agree on this one. Big Sur is a catastrophic release. It really shows the impact that working remote can have on a major software development project. I am kind of surprised how there are not more major issues, but in areas where it requires hardware testing, macOS has really gone into the tank. The way the security over-rides the driver installation project is fundamentally broken. Apple needs to fix this ASAP, not rush into another major OS release with yet another major driver change, which is what appears to be the case. We will find out in June what Apple has planned, but I would rather see bug fixes than 100 new features.

Its absurd to have so many issues installing drivers. SoftRAID installs and functions mostly the same as it has for 18 years. What has changed? Big Sur has a hard time even updating the extensions cache reliably.

If you quit and immediately relaunch, does the app run? We may have to walk you through a uninstall.

@softraid-support that may well be but it's only with OWC Raids and I had the same issue with a MacPro which I then replaced with a 20205K iMac

I now have 2 x Drobos and a Promise raid and they all work fine.

And yes I did months of trouble shooting with countless clean re-installes before selling the OWC enclosures I only bought a few month before