Notifications
Clear all

SoftRAID Volume Locked error

Page 8 / 13
(@ionblue)
Active Member Customer
Posted by: @softraid-support

@ionblue

What if you plug the hub into the "high power" port on the dock? does that make any difference?

Why would that make a difference? The drive isn’t connected to the hub or dock. It’s directly connected. And I don’t have anything connected to the hub that needs extra power. 

ReplyQuote
Posted : 13/06/2021 8:39 am
(@ionblue)
Active Member Customer

@softraid-support Forgot to mention that it hasn’t happened since I disabled sleep on the MBP. Already had the HD sleep option disabled.

ReplyQuote
Posted : 13/06/2021 8:57 am
(@softraid-support)
Member Admin

@ionblue

Just testing. MacOS is weird right now. Connections are going at incredibly high speeds. Small amounts of interference can be an issue, so trying different connections doesn't hurt. I have seen weirder things than moving the connection to stop a volume eject issue.

ReplyQuote
Posted : 13/06/2021 10:21 am
(@srogers1)
New Member

I had the same thing happen today on wake from sleep. Currently volumes don't mount automatically. I'm mounting them manually by running: 

sudo kmutil load -p /Library/Extensions/SoftRAID.kext

My Thunderbay is connected directly to the iMac. The only other devices are USB.  Support file attached. My logs look just like all the others, with error E00002E4.

Hopefully, you can figure this out. (BTW - switched to Thunderbay from Drobo, because Drobo is clearly going off a cliff, even though I really liked the zero-config, zero-work setup)

iMac Pro (2017) - macOS 11.4, 3 GHz 10-Core Intel Xeon W, 64 GBf

Thanks,
SR

 

This post was modified 1 year ago 2 times by srogers1
ReplyQuote
Posted : 14/07/2021 2:30 pm
(@softraid-support)
Member Admin

@srogers1

We do not have a workaround for this Big Sur bug at this time, we probably have to wait for the macOS fix, which may be an a future Big Sur release or in Monterey, but we were told it will be fixed.

Note:
The "up" key in terminal will show prior commands, so next time you need to manually load SoftRAID, just hit the up arrow in the terminal window, and this command will be available without having to type it.

ReplyQuote
Posted : 14/07/2021 10:56 pm
(@hedware)
Active Member

@mygiguser Yes this problem has happened all of a sudden. I am now getting this problem frequently. I've changed nothing except upgrade to latest version of SoftRAID. This problem is only happening with SoftRAID configured OWC enclosures and drives - not with other drives.

I have looked through this series of posts and I am getting rather tired of SoftRAID saying it is everybody else is at fault.

I am not interested in getting into the technical aspects - I dont have that time - SoftRAID needs to clearly state whether its software is ever going to work without this locking up problem happening. If not then I want a refund.

ReplyQuote
Posted : 12/08/2021 8:31 am
(@softraid-support)
Member Admin

@hedware

If you are referring to the volume locking error, SoftRAID is proactively preventing data corruption.

For example when your computer is asleep, if the drives spin up during this cycle, that is one of the main triggers, as this means the disks were re-powered up (macOS generates new disk identifiers for the disks). Renumbering disks while the system sleeps is clearly an issue. Its primarily a Thunderbolt issue, as if a Thunderbolt chip "crashes" or resets, all disks connected through it are powered off instantly, triggering this issue.

We used to see data corruption from this bug, which have disappeared since we implemented this test at waking from sleep. So while unfortunate to have a configuration that has this problem, volume locking is actually protecting your data.

Are you using a dock? Thunderbolt Monitor? or is this a simple one computer/one enclosure scenario?

You do not see this with other disks, as they are single disk solutions. If they power off during sleep, it does not cause data to be written to different disks in the RAID volume out of order.

 

ReplyQuote
Posted : 12/08/2021 9:51 am
(@hedware)
Active Member

@softraid-support Thank you for your reply. I have a string of multi-disk enclosures, some in a Thunderbolt daisy directly from my computer. I have monitoring equipment that serially connects to the computer and I have hourly remote backups. These cause the computer to wake up automatically.

This problem has only just started happening in the last few weeks. Once I have unlocked the locked computer, my test apps do not find any problems with drives, volumes, etc. SoftRAID draws the same conclusion.

It is not much of a test if it causes this problem. From the responses to this issue, I am none the clearer. It would help if you could give a clear and explicit problem statement. It would also help if work-rounds could be suggested or even implemented particularly if Apple is not interested in helping third-party problems. I appreciate that Apple software leaves a bit to be desired. Otherwise, how about giving an option to disable this testing by SoftRAID.

I gather from doing a Google search that this problem is more common than SoftRAID imagines.

ReplyQuote
Posted : 17/08/2021 1:27 am
(@softraid-support)
Member Admin

@hedware

The problem is macOS assigns semi-random disk numbers to each drive at startup/connection. These are "permanent" and what the OS uses to keep track of the disks.

the problem is if during sleep, the thunderbolt bus resets, which causes the disks to get different numbers. Then If macOS tries to write out to the disks, data can go to the wrong disk.

Its not much of a problem with single disk volumes. But with RAID, it means sending data to the wrong disk, which would currupt data.

 

You cannot (do not want to) turn off this check, as it prevents your data from being damaged.

 

If this only just started, there might be a cable that is causing this at sleep. Or a dock. You might be able to isolate this by moving connections around, or looking for any loose connections. If you find all devices on one chain are doing this, then you can look in that chain only. Try process of elimination. Its a frustrating bug.

We talked with Apple about a workaround and might be able to create a solution that lets SoftRAID figure out the correct disk numbers, without compromising data integrity or performance.

 

ReplyQuote
Posted : 17/08/2021 10:17 am
(@angryghandi)
Active Member

I am also experiencing this issue so routinely it has made using my computer with the enclosure attached unusable.

My enclosure is plugged in to a USB-C port on the left hand side of the computer along with an LG Ultrafine 5K display. I will try moving the enclosure to a port on the right hand side as you have advised others in previous messages.

In the mean time, I am including a Tech Support Report to see if you can find any other issues and have any additional recommendations to resolve this issue.

 

ReplyQuote
Posted : 14/09/2021 7:45 pm
(@softraid-support)
Member Admin

@angryghandi

Curious what happens when you separate the display from the Thunderbay, as you indicate. No I do not see anything obvious.

What is with the unused 3TB drive? Is that essentially a "warm spare"?

ReplyQuote
Posted : 15/09/2021 1:29 am
(@hedware)
Active Member

@softraid-support I thank you for your response. I am still getting these problems which as I have said are a relatively recent experience.

Since your reply I have been keeping notes and the problem only seems to happen to a backup OWC 4 bay enclosure with 4 drives set at Raid 5. This enclosure is attached to another primary OWC 4 bay enclosure which in turn is attached to the iMac.

Both enclosures and their drives have been checked with Disk Warrior, SoftRaid and other test apps - the result is that everything is in order - zero faults. There are a couple of other drives formatted using SoftRaid and attached directly to the iMac that do not exhibit this problem

This problem is happening with the enclosure at the end of the daisy.

If it is an Apple problem as SoftRAID alleges, then why does this problem not happen to the enclosure that is next in line to the iMac? 

Can you proffer any ideas why it is the enclosure at the end of the daisy?

Is it an OWC problem and not a SoftRAID problem?

 

ReplyQuote
Posted : 18/09/2021 7:12 am
(@softraid-support)
Member Admin

@hedware

The source of this problem is two fold:

1. Thunderbolt is a very high frequency bus, in the gigahertz range. Same as cell phones, and other high frequency devices, and is susceptible to interference. (the chips generally "reset" when this happens)

The Thunderbolt chips themselves can crash from heat, interference and probably other reasons.

2. Thunderbolt was "designed" to instantly power off devices when disconnected. So any interruption in signal, for any reason at all, will instantly cut power to all connected devices. Disk drives, as in this example.

The combination of these two issues trigger the disk eject alerts.

That is why resetting or replacing cables can help, why additional devices, even monitors and USB hugs, thunderbolt docks, multiple enclsoures, etc can cause this.

When the chips reset, it is like pulling the cable and replacing it. When MacOS sees the drives reconnect, they are assigned "random" disk identifiers, which generally do not match their previous numbers, in the IO registry. If SoftRAID driver did not detect this, data would be written to the wrong drive, causing corruption.

 

Hope this helps you understand the causes of this. Maybe you have one bad cable, or the enclosure is resetting, or it is not connected tightly.

This post was modified 1 year ago by SoftRAID Support
ReplyQuote
Posted : 18/09/2021 1:46 pm
(@steve223)
Estimable Member Customer

@softraid-support strange thing is that neither my Drobo D8 nor my Promise Raid with Thunderbolt 4 has this issues. 

ReplyQuote
Posted : 18/09/2021 9:47 pm
(@softraid-support)
Member Admin

@steve223

You do not know if they do, probably. As they do not have the "detector" that SoftRAID does. In addition, they only present a single disk to macOS, where SoftRAID is managing 4 disks in an enclosure.

ReplyQuote
Posted : 18/09/2021 11:28 pm
Page 8 / 13
Share:
close
open