I recently upgraded to macOS 10.15.6. Now if I leave VM Ware Fusion running while leaving my mac for some time, e.g. - for a night, the system crashes. This did not happen with the older macOS 10.15.5 version.
VM Ware Fusion version: Professional Version 11.5.5 (16269456)
macOS Catalina version: 10.15.6 (19G73)
I didn't need to re-install anything.
Everything (applications, data etc) was intact (but it's always prudent to ensure you have a backup for plan B).
Dave.
Finally i use time machine to come back in 10.15.5 .... waiting for a new release of Apple ... or with Big Sur
I'm no longer having crash reports pop up when I reboot after a crash. Maybe Apple is tired of getting them.
Does anyone know how to get this information on a Mac?
As I understood, popup report is generated when the system reboots itself. If you rebooted it manually, the report will not be generated.
Same here. It's quite annoying to spend time & effort on this especially that I have far more important things to do like deadlines to meet. I moved my work related stuff to my "in case of emergency" mac (wasn't updated, I was suspicious of 10.15.6). It would be nice if all this hassle could be invoiced on Apple 😕 Great job on pinpointing the root cause.
i just got off phone with apple tech support who called me because of my posts on this issue on apple discussion boards.
single-minded focus, trying to get me to induce the crash so a diagnostic report could be generated, even after i pointed to various professional opinions who thought the problem could not be resolved at user level. was not interested in bug having been reported with steps to reproduce.
in one sense i'm glad they scan their boards for crash cases. in another sense i'm furious the rep gave me the feeling it was my fault, and that i had to induce the ugly crash i've spent a couple weeks trying to work around, or maybe it wouldn't get fixed.
Too bad Darius can't share the radar ticket number...'hey, go read this'
I had the same thing when I called the developer program trying to get the installer....oh, wow, never heard of that.
I too was pulling my hair out with these crashes. I was pretty sure it was 10.15.6 because it was stable prior to this update. Thanks to the MDS tip, I was able to download 10.15.5, create an USB installer from it, and do a fresh install of 10.15.5
I will be visiting this thread periodically to figure out when its safe to do another upgrade again.
thanks,
Kiriakos
It happened again. Same as before as I was typing something in a guest Windows machine, the system started behaving weirdly. I was able to record a video on my iPhone. I noticed that not all services went down and I was able to type "sudo shutdown -r now" in terminal, which worked correctly, but VMWare did not shutdown properly and after the reboot, Windows was restarted rather then woken up.
It looks like it might be related with windows VMs, I run linux VMs under 10.15.6 with no crashes. On my other system with windows VMs, constant crashes!
I submitted a report to Apple with a video of macOS going berserk. Hopefully they'll get back to it.
I run Linux VMs 95% of the time, and experience this panic/reboot issue regularly, so I do not think it is related specifically to Windows VMs.
it's not related to guest os, it's related to how much the guest system's behavior pokes the bear inside 10.15.6.
windows pokes it hard just by breathing.
i have a xubuntu guest, very low impact in general, but i can make the bear dance like a puppet by increasing & decreasing network file i/o. leak faster! now slower! now faster!
this is why i have istat menus on my host menu bar now, showing only memory pressure % in bold, purple text. the guest os is gonna do what it wants. the only reliable way i've found to tell how my 10.15.6 host is reacting is by watching how hard it's having to work managing memory. for me, a crash is near certain once memory pressure reaches 65% so i try to restart before it reaches 60%, and i have istat alerts set to warn me as mem pressure climbs past 50%. last crash i had was five days ago. restarting host twice a day.
ps. istat also has history, which has been super helpful figuring out what to do. the last crash i experienced was thursday evening.
memory pressure basically shows the wired memory growth. the downtime on saturday was weekly maintenance.
friday 2p is more important though. the first peak was when i got a warning & i quit some background apps & logged out a user who was done for the day. the second peak warning was when i shut down the vm & finished what else i was doing before restarting.
the peak on monday was when i was deciding whether or not to let the apple tech persuade me to induce the crash. it was looming.
I know it was somewhat shutdown to use the old kernel with the 10.5.6 system, but has anyone tried transplanting *just* the 10.5.5 Sandbox.kext?
I’m guessing it’s too tightly integrated but it might work.
As a side note, I can reproduce this on my ancient tried and tested VirtualBox 5.X setup. I haven’t had any panics but I can definitely see the leak with zprint.
Is there a way to free the kalloc.32 memory that is not being used without rebooting?
I'm using virtualbox and been experiencing the same issue.
After several crashes (2-3 each day), I too figured out the kernel memory leak problem since 10.15.6 update.
Now that I'm aware that the com.apple.security.sandbox Kext is actually causing the memory leak, I'm eager to find a workaround.
As a workaround I did copy the /System/Library/Extensions/Sandbox.kext from 10.15.5 to my Macbook running 10.15.6.
For now it's running without any problems and the memory leak seems to be resolved.
I will keep my Windows 10 guest running and post any updates.
Well awesome. I'm glad I'm not completely crazy for suggesting the idea of transplanting the Sandbox.kext.
I wonder if it'll cause any negative effects.
Sources for 10.15.6 are available for download here Maybe someone could have a look inside and fix it?
If it truly is Sandbox.kext (which it looks like it is) then that module is closed source.
If we somehow could find and correct the code ourselves, we'd run into the issue of our version not being signed correctly.
I see. Where can I download the previous version of Sandbox.kext?