Frequent kernel panic in IONVMeController.cpp

UPDATE: They finally let me return the drive. It's a shame because this thing is SO fast but I can't take a chance with my MacBook since I use it for work. I still only experienced the Kernel Panic the one time but the standby issue is prevalent unless I disable standby via Terminal.

I would suggest that for anyone looking at these drives that they only use them for desktop systems that are going to be on all the time. If you're like me and you pack up your system to go home every evening and leave the system sleeping on battery for several hours, you're going to run into the reboot issue when resuming from standby. And if your system is critical to doing your day to day work and you don't have a backup you can use, then you have to worry whether these kernel panics are indicative of an unsupported configuration.

Best of luck to anyone else trying the drive. I hope you have better luck than I.

I had a simlar experience. Kernel panics (4 or 5 a day) under 10.13.3 with new Aura Pro X 2TB - clean install on a Mac Pro 2013. It seemed to happen most when the I/O was maxed out downloading apps from app store and uploading files to Dropbox.

I went on the OWC chat line (they did ask me if I had tried to troubleshoot the panic log with an implication that they would help in this). However, in the 1st instance they suggested resetting SMC and PRAM. Resetting the SMC made no diference (in fact I had a kernel panic on reboot), but resetting the the PRAM seems to have fixed it (fingers crossed 48 hours of intensive I/O and no panic). To confuse matters I also disabled "put hard disks to sleep when possible" soon after so it could be that !!!

OWC advice:

Chas B (08/03/2018, 21:02:36): SMC for the machine is just pulling the power cable for about 20-30 seconds and then plugging back in.

Chas B (08/03/2018, 21:02:38): PRAM Steps:

1. Power off machine.

2. Hold Command + Option + P + R

3. Power machine on with the keys held.

4. Hold until your machine reboots 4 times, then let go.

5. Boot normally

EDIT

I have since read Rasputins426's post. Maybe it was unchecking the "put hard disks to sleep when possible"

I'm also seeing this kernel panic under 10.13.2, with the 1TB version of the Aura Pro X and on the same late-2013 MBP. It happened every night (and sometimes during the day) until I disabled standby from the terminal ("sudo pmset standby 0"). I'd also previously turned off Power Nap and wake for network access in an attempt to deal with Sleep Wake Failure issues that I saw under installing this SSD and 10.31.1. It's gotten better with these changes, but I still get this kernel panic every week or so.

I also had the UI lag/CPU issue, but this was only while the disk was being encrypted after I turned on File Vault (3 long painful days there).

I have a support request out to OWC, but it's been a while and I haven't heard back from them (odd, since it appears that they know about the problem). It's not sounding like I'm going to get much satisfaction there. Sigh...

The kernel panics may be related to encrypting the drive- that’s about when I experienced it as well.

The double reboot when coming out of standby issue (you may have thought it was a KP but at least in my experience I never saw a KP) is something I had verified as a known issue by at least one OWC employee and disabling standby through terminal is the only fix. I read in another forum that this is because the NVMe controller is not available fast enough for the OS to resume and so it forces a restart until it detects the drive. That is such a low level issue that, as told to me by an OWC support agent, OWC would have to work with Apple on a fix. And we all know how incentivized Apple is to force an OS change to support 3rd party internal hardware...

I’m glad OWC finally let me return the drive. Yes, I miss the extra space and the performance but what other issues could be there? I just wish they would remove that “guaranteed compatibility” claim from the product page. I find it VERY telling that 1) there haven’t been any reviews from the usual Mac blogs and 2) the two user reviews I’ve tried to post to their site haven’t been published.

My kernel panics are all related to the sleep/wake/standby issue, not heavy I/O. They end up as panics in this same file but on a different path. See below. I'm not saying that heavy I/O isn't also an issue. It's just not something I do much on this machine (beyond the encryption, which was on an earlier patch).

I was counting on Apple's shiny new APFS to protect my files during these panics but... no. I had a file corruption issue after one of them. I don't remember ever having that happen with the old file system. So here's another vote for returning this SSD. Unfortunately for various reasons I let this languish beyond the return period so I don't know if that's an option for me. I'll have to have that discussion about "guaranteed compatibility" with them.

I'd wondered about the lack of reviews for this SSD on OWC's site. I suppose a generous interpretation is that they've been too busy to approve them. They were also very slow to respond to requests for information about using the old Aura drive under High Sierra. Still, it's not a good look.

Sun Jan 14 00:31:55 2018

*** Panic Report ***

panic(cpu 0 caller 0xffffff7f846bf82a): nvme: "IONVMeController::HandleControllerPowerOff, ShutdownNVMe() completed with status = 0xe00002c9\n"@/BuildRoot/Library/Caches/com.apple.xbs/Sources/IONVMeFamily/IONV MeFamily-356.30.6/IONVMeController.cpp:2011

Backtrace (CPU 0), Frame : Return Address

0xffffff81eee33bc0 : 0xffffff80012505f6

0xffffff81eee33c10 : 0xffffff800137d604

0xffffff81eee33c50 : 0xffffff800136f0f9

0xffffff81eee33cd0 : 0xffffff8001202120

0xffffff81eee33cf0 : 0xffffff800125002c

0xffffff81eee33e20 : 0xffffff800124fdac

0xffffff81eee33e80 : 0xffffff7f846bf82a

0xffffff81eee33ea0 : 0xffffff7f846bb740

0xffffff81eee33ed0 : 0xffffff8001289e74

0xffffff81eee33f40 : 0xffffff8001289965

0xffffff81eee33fa0 : 0xffffff80012014f7

Kernel Extensions in backtrace:

com.apple.iokit.IONVMeFamily(2.1)[8C2350A2-A812-3B8C-8DE1-3A6911793417]@0xfffff f7f846b0000->0xffffff7f846ecfff

dependency: com.apple.driver.AppleMobileFileIntegrity(1.0.5)[58669FC2-CC90-3594-AD69-DB89B9 23FD20]@0xffffff7f82470000

dependency: com.apple.iokit.IOPCIFamily(2.9)[C08F7FC1-78A4-3A1B-BFE2-C07080CF2048]@0xffffff 7f81a94000

dependency: com.apple.driver.AppleEFINVRAM(2.1)[79C21A42-B04F-314F-BA88-F221F3D54094]@0xfff fff7f8275e000

dependency: com.apple.iokit.IOStorageFamily(2.1)[B0FCE898-1542-34C0-B845-46FAE81DD9E6]@0xff ffff7f81dc9000

dependency: com.apple.iokit.IOReportFamily(31)[D2F2FBDF-4EE4-38BA-99F5-B699F886F413]@0xffff ff7f82784000

...

Just a quick update ... since I recall making another change which was to my Energy Saver settings. Since I wanted FileVault Encryption to run a bit faster I turned off all energy saver settings (only box checked is to prevent computer from sleeping when display is off). I've also reconnected all peripherals and set up is the same. So far so good. I've also been through multiple boot cycles and I can't reproduce the problem. It used to happen on every single shutdown. Very strange...

Another update that I hope will help someone else avoid the pain that we've been through with this...

I spoke with someone at OWC who seemed knowledgeable and he told me that they are not aware of any issues and immediately sent me a replacement. I went through every troubleshooting step I could think of and the OWC Tech Support rep was confident that it was an issue with the OWC Aura Pro X SSD that I received.

I have not installed it yet but I will report back when I do. In the meantime I have some clues about what might be causing the issue. My hypothesis is that it is heat related and not about load, I/O or boot/shutdown. Here are the clues:

1. The Auro Pro X SSD comes with a USB 3.0 enclosure that OWC states should only be used with the original Apple SSD. I dug into that a bit with OWC and it's because the enclosure will not dissipate enough heat to allow the OWC SSD to function without risk of damage. If it needs to be used with the OWC SSD then it should be run with a fan blowing air across it and with the enclosure open. I hope that this is not a design defect but it is possible that when installed in the Mac Pro or MacBook Pro that heat is still an issue.

2. After a long day of relatively light usage I started to experience issues as I got near the end of the day... stuttering audio and video, cursor/mouse lagging, etc. The Aura SSD had been under much heavier load earlier in the day with absolutely no problems.

I have nothing to add other than I just experienced this exact sequence of events on my newly installed Aura Pro X. I'm also seeing an issue where the computer will reboot when you let the system go into standby / deep sleep / hibernate (sleep for 3+ hours while off AC power). I already called OWC about that issue and they said they're working with Apple on a fix but no ETA...

I'm really starting to regret purchasing this drive from OWC. I also bought a demo model to save a little money and they're refusing to allow me to return it for a refund.

I’m sorry to hear. I would call them and discuss the standby issue. I would call their tech support team, not customer service as it was tech support that approved my return. I also didn’t have to escalate or ask for a manager- I was just calm and asked for help. They may offer to do an advance replacement of the drive or even have the drive sent in for testing if you’re up for that but I don’t think either option will be of any benefit as this issue appears to be systemic and borne from a limitation in how macOS reads from NVMe drives.

I think unless they tell you there’s a fix in development and they give you an ETA, they should allow you to return it for a refund. It’s possible they will push back but they can’t leave you with a defective piece of equipment. I worked hardware engineering escalations for another computer company for years and at least for us in this same sort of situation, we did the right thing. And since they allowed me to return or of policy (I bought a non-returnable demo unit) I’m confident OWC will help you.

Apologies for not replying sooner, for some reason I was not subscribed to this thread.

Their chat support is abysmal for this type of discussion. l called in and although they didn't have much more information, I was at least taken seriously. That chat log you posted above is dismissive and unacceptable, in my opinion.

All that said haha, I called Friday, had a new agent take ownership of my case who requested crash logs (which sadly did NOT reference the kernel panic I experienced - still just one KP so far thank god), and well, that's been it so far. I emailed her yesterday requesting an update and I'm hoping I didn't get a reply because she doesn't work on Mondays. Having worked in tech support call centers before, I know their schedules can be erratic.

Either way, I think your best course of action is to return the drive to the vendor. I wish I could but since I bought a Demo item directly from OWC, they are refusing to let me return it. I even requested just store credit so I could buy something else and they still won't let me. I've never in my career heard of a company that has a verified issue (the standby issue) with no fix for an ETA not allowing an exception for return.

Also, I find it very funny they were pushing you to talk to Apple when my support rep told me they were working directly with Apple on a fix lol.

PS- I left a review on the 480GB SSD on their website that they haven't 'approved.' Very sketchy behavior for OWC.

Just adding my voice to the chorus. This time with a Mac Pro 2013 and 10.13.2. Problems seem to have stopped after I turned ON FileVault and disconnected all drives and peripherals. I'll run like this for a while to see if they're related or if it was just a coincidence.