Since installing Catalina yesterday, multiple crashes from userspace watchdog timeout

So a work around (until Apple realizes and fixes this problem) that actually worked for me was to use an app to prevent Mac from going to sleep. I found Amphetamine really useful with bunch of really awesome feature and it's totally free.

Note: I tried Mac's built-in way to prevent it from sleeping automatically (when connected to power), but it didn't work!

tm1uet wrote:

For those of you who are experiencing this issue and using a CalDigit TS3 plus hub there are good news. Last January 6th 2020 CalDigit released a firmware update for this device that seems to fix the issue on my MacBook Pro 15" with Radeon Pro Vega 20 + 2 external monitors under macOS 10.15.3. Since then I haven't experienced more reboots when the screensaver was on.

This is the link to the update:

https://www.caldigit.com/updating-the-ts3-plus-firmware-via-macos-10-15/

Did it stick? How long have you been stable since the update? 🙂

Same here, also connected to external monitor (sometimes one with origin apple adapter or two monitors with OWC dual HDMI adapter). Also happens that I cannot shut down or restart my Mac properly. Mac crashes and reboots automatically, with a previous cooler fan-boost. I have a 6-core 16inch MacBook Pro.

panic(cpu 6 caller 0xffffff7f9939dad5): userspace watchdog timeout: no successful checkins from com.apple.logd in 120 seconds

service returned not alive with context : unresponsive dispatch queue(s): com.apple.firehose.drain-mem

service: com.apple.logd, total successful checkins since load (6920 seconds ago): 681, last successful checkin: 120 seconds ago

service: com.apple.WindowServer, total successful checkins since load (6890 seconds ago): 689, last successful checkin: 0 seconds ago

Backtrace (CPU 6), Frame : Return Address

0xffffff8216adb820 : 0xffffff801893bb2b

0xffffff8216adb870 : 0xffffff8018a734d5

0xffffff8216adb8b0 : 0xffffff8018a64f4e

0xffffff8216adb900 : 0xffffff80188e2a40

0xffffff8216adb920 : 0xffffff801893b217

0xffffff8216adba20 : 0xffffff801893b5fb

0xffffff8216adba70 : 0xffffff80190d2b25

0xffffff8216adbae0 : 0xffffff7f9939dad5

0xffffff8216adbaf0 : 0xffffff7f9939d7e6

0xffffff8216adbb10 : 0xffffff801906739b

0xffffff8216adbb60 : 0xffffff8019070443

0xffffff8216adbca0 : 0xffffff8018a22d12

0xffffff8216adbdb0 : 0xffffff80189419d8

0xffffff8216adbe10 : 0xffffff8018918635

0xffffff8216adbe70 : 0xffffff801892f0e5

0xffffff8216adbf00 : 0xffffff8018a4b575

0xffffff8216adbfa0 : 0xffffff80188e3226

Good day people,

This post should be considered by people who can keep a sane machine with an SSH connection on the emotional machine (that reboots without notice).

Observations

------------------

Shortly after boot, WindowServer is visible in the process list

ps -ef | grep -i window

   88   213    1   0  2:30PM ??      0:06.04 /System/Library/PrivateFrameworks/SkyLight.framework/Resources/WindowServer -daemon

and shows up in launchctl as well:

sudo launchctl list | grep -i window

213 0 com.apple.WindowServer

However, as the screen is frozen, WindowServer is not to be seen anymore in the process list, and has status "-6" in launchctl.

Suggestions

-----------------

1. For the people inclined to maintain an SSH connection on your emotional machine (mine is a MacPro 2013), could you please confirm you observe the same thing before freeze, and during freeze.
2. I have attempted to restart WindowServer with launchctl but to no avail. Anyone had success trying that?
3. I have attempted to restart WindowServer without "-daemon" argument, and it failed indeed. Now I know better I'll try when it freezes again. (I will create a little shell script to do that for me...). Please report your attempts / failures / success in doing so.

Looking forward to get confirmations and counter examples !!

We need to find a work around that will allow our long winded processes to complete without being reset in mid air by a pesky bug in the OS or a GPU driver.

Just had the hang happen again.

In my case WindowServer still seemed to be up and running.

Hopefully everyone having this problem has opened a support ticket with Apple to try to get their attention. This really should be a high priority for them...

The latest upgrade to 10.15.4 fixed my watchdog timeout errors. (Finally!)

FYI My system is a 2018 15" MacBook Pro 2.9 GHz i9 w 32 GB ram & a Radeon Pro Vega 20. I use it in clamshell mode with 2 adaptors driving three external 24" monitors.

The only fix I'd previously found to stop the error was to leave a QuickTime movie continuously looping in the background.

Catalina was stable for quite a while. 15.4 seems to be the update that broke it. The computer is so unresponsive now that I can’t even access any peripherals to see if I still have my bootable backup. I usually delete them after a new OS proves stable for 3 months. In the past, every OS update in the system X era has made my computers faster and more stable. Until now. From what I’ve read on threads, you can’t use time machine to downgrade from Catalina without a lot of data loss. Is that not the case?

Nothing is perfect, but my computing experience has been suck to miserable since December 2019 when I upgraded. I am reluctant to bite the bullet an do a clean install because I understand the work involved getting me whole again. I tarted down this path (clean install) today and saw that the problem is still widespread and not fully understood. I am embarking on that path today. Not with Catalina, but Mojavie.

... not to mention the increase in my profanity... :^(

[Edited by Moderator]

Hi All

(Mac Pro Late 2013, D300 Graphics, 4k Displayport screen, Latest OSX)

To feedback to the last two posts;

@apple-support-today > caffeinate -u -d

Didn't help. Even running it as sudo didnt

@jmsunico > Disable power nap & call notifications

Didn't help.

Just cleared SMC/VRAM (again), disconnected everything disconnect-able leaving screen on displayport, keyboard on usb and mouse on keyboard's usb. Caffeinate running.

Lets just say that it hasn't improved and this is now the second time I'm typing this post. 35 mins since last forced reboot. Mail/Excel/Safari open + usual background junk on toolbar.

@jmsunico > My hypotheses is that the problem is really deep and it has to do with memory buffers affecting at least hard disks and video memory

On the whole I would agree. As frequency of hangs is certainly higher when working with video (play/edit). However the hang just now was while typing the previous version of this reply. Mail was open to get the thread link, Excel because it recovered the spreadsheet I was working on, terminal to caffeinate and safari to type here ... not exactly a busy system.

"...Opening a small .psd file on google drive makes photoshop freeze for 10 seconds, same happens when saving.

..."

What happens if you download the file first then open it? You have to include the possibility that PS and GDFS have a problem working together in the Catalina environment.

What takes 10 seconds? The download or the opening?

Does it cause a crash?

I continually get these same panic codes and restarts. My restarts happen sometime while my MBP is idle, but most of the time it is when I am ripgrep'ing throw my file system and I attempt to Control-C out of the search my mouse and everything else becomes unresponsive. When this happens the panic results and a restart happens.

RogerDavis wrote:

Hi Bob,

I do not use Time Machine. I have an attached HooToo USB hub with (i) power input, (ii) external mouse and sometimes (iii) printer connection. I also have a Belkin USB-C ethernet dongle plugged directly into the Mac, and sometimes a portable USB hard drive.

I never experienced this problem until I upgraded from 10.15.4 to 10.15.5 about ten days ago. I was using the same hardware setup for about 3 months prior to that, i.e. 2019 PowerBook with above doodads running in clamshell mode connected to an external Philips monitor. The thing went to sleep every night and woke up happily the next day for 90 days in a row. Since the 10.15.5 update it reliably hangs and reboots every single time it tries to wake up (until I started using the caffeinate workaround to keep it from sleeping).

Hope this info helps.

Well, it says the Time Machine software is not the actual cause, but if the person at work was using a USB device, then maybe the common element are external devices, and a change to the macOS Catalina USB device drivers.

Historically (and I'm going back to some very early Mac OS X versions), there have been times perfectly working devices (or in one case 3rd party RAM) that started causing the system to panic after a major operating system upgrade. Now this could be because the device was not fully in spec, but the previous software did not exercise it in such a way that the spec flaws caused any problems. Or this could be changed macOS code that is not taking into account some real-world situations, and we get this watchdog timeout.

I do not know, but as I seem to like looking at kernel panic reports, I would like to either have watchdog timer magically go away (because I can never figure out from the panic report why it happened), or find a common root cause that I can give useful diagnostic suggests to users experiencing it, that might lead to them not having it any more.

So thank you for your feedback.

PS. As with most users in these forums, I'm just another user. In my day job I do work inside the kernel of other operating systems, which is why I seem to like looking at panic reports.

I am actually downloading 10.15.6 now, and will report back asap. I have been struggling with 1-3 daily crashes for weeks now, and as a software developer, it always hits when I have 50 things open and in-progress. I have one external monitor directly connected to my MBP (15", 2019) using a DVI->USB-C adapter, and the other external monitor is connected (via HDMI) to a Lenovo USB-C docking station (just for reference).

Catalina, unfortunately, has been nothing but a headache since I installed it. I expect more from a $2K+ laptop (and the company behind it)...

-- Kevin

Well, my success was short-lived. I had been working for about 30 minutes today, and got the spinning wheel of death (and had to force reboot). And, interestingly, I am not using ANY external monitors. So that somewhat rules out that connection (at least in my scenario). Something else still must be happening to cause these crashes. I did have Chrome browser open, and a news headline video was playing in a non-active page (not saying that is relevant, but Chrome browser was the ONLY app open at the time of the freeze/crash).

Never thought I would be at the point where I was wishing the company had given me a Windows laptop vs. a MacBook Pro...

My success was also short-lived. I also had been working for about 30 minutes today, fired up VMWare Fusion and had everything freeze (I don't get the spinning wheel, just a complete freeze-up and finally an automatic reboot after 1 - 2 minutes). I was using my 3 monitor setup. I was not using the Chrome browser (Chrome browser is not even running, not turned on).

So I disconnected my third monitor and turned off Power Nap. Not likely to help but we'll see.

And yes, I'm now using my company-issued Windows laptop for mission critical apps (like MS Teams). The Windows laptop never crashes. Very sad...