Where does the bug appear (feature/product)?
Grok Bot
Describe the Bug
Hi, my Grok Bot computer is being paused from outside the VM, overnight and repeatedly during the day. Nothing inside the box crashes. I collected evidence before touching anything; I have not run Reset, Recover or Update.
Plan / app: SuperGrok Plus, Mac desktop app, 12 Bots on one computer
Computer: grok-bot-vm-416470035, last boot 2026-10-19 08:56 UTC, no reboot since
Resources are fine: 15 GiB RAM with no OOM or hung-task messages in dmesg and cgroup memory.events all 0; disk 125 GB at 14%; load near 0
Main pause: 2026-10-20 17:18 UTC → 2026-10-21 02:51 UTC (about 9h32m). Three independent processes stopped logging within seconds of each other and resumed within about 2 minutes of each other:
WhatsApp sync daemon: last line 17:18:24, next 02:53:20 UTC
a 2-minute board watchdog: last tick 17:18:18, next 02:52:07 UTC
rclone mount: silent from about 17:15, resumed 02:50:44 UTC with context canceled / Device or resource busy errors
Our monotonic-vs-wall-clock check recorded “frozen 34308s”. Cron stopped ticking for the whole window.
Shorter pauses (UTC): 10-20 15:01 (56 min), 17:01 (117 min); 10-21 02:23–04:18 (57 min), 05:08 (20 min), 05:45 (20 min), 06:12 (14 min), 06:51 (47 min), 07:40 (9 min), 07:57 (4 min).
My Mac stayed awake all night (pmset shows no sleep), so this isn’t the desktop app disconnecting.
Questions:
Is this computer being suspended or live-migrated by the host? Is it related to the 10-17 account/computer move?
Can it be kept running 24/7? Bots need it live overnight.
Is there anything I should do (or avoid) on my side?
Thanks.
Steps to Reproduce
It happens automatically and without warning.
Expected Behavior
Computer and associated jobs should continue running all night.
Operating System
MacOS
Version Information
Version: 0.72.0
Built: 2026-10-18T04:33:44.546Z
Release Track: stable
OS: darwin
Does this stop you from using Cursor
Sometimes - I can sometimes use Cursor