I have been having this problem for over a year and have searched for clues to its resolution and found almost nothing. The problem is that after a time (could be days or weeks), cmd.exe processes that I started manually go to background and cannot be restored to foreground. Is there a way to bring these back to foreground without shutting them down?
Used procexp, taskmgr, taskkill, sfc scannow, restart Windows Explorer, etc. No joy. They can't bring them back, even if killed and restarted. The processes themselves usually continue to run correctly, but the machine may get really slow. What seems to be happening, one aspect, is that the registry/hive entry for that process label gets corrupted (memory issue?). (note: I always give my important processes unique labels, for example [start "DATAABC" python.exe something.py], then change its properties as needed.) Once the problem happens, whenever I start DATAABC, it disappears and cannot be brought to the foreground. I have to regedit, find DATAABC, delete the key, then re-start DATAABC, and reset its parameters. Annoying and somewhat risky with regedit.
The processes run on a dozen PCs and exhibit nearly the same behavior. Sometimes it resets my network connection, but I can recover from that without rebooting.
I use things like "setlocal enableextensions enabledelayedexpansion" with a for loop, tasklist, robocopy, python 3.8, and move tens of thousands of files a day on each machine. I use the machines pretty intensely, but CPU utilization seems to be < 50% and physical memory usage stays below 50% on 16GB according to taskmgr. SQLite3 eventually buffers the tables and indexes into "unused" memory (about 4GB) and dramatically speeds up at that point. I run four instances. A fifth bogs down the machine.
Is there a way to restore these processes to foreground without shutting them down and doing the regedit dance? Note that if started from schedtasks rather than manually, they are always "Run only when user is logged on". I suppose I could export their registry entries when they are good and then import them when they go bad, but that seems kludgy, and I haven't tried it. I cannot replicate the problem at will. It just happens over time.
Thanks for your consideration of the issue. I am sure there are other HW/SW issues that lead to this, but nothing shows up in device manager or in the system logs/events, so it doesn't seem worth my time to attempt to diagnose. I can live with that, barely. Using various brands of SSD for the system drive and Seagate ST4000s for some of the data drives and Seagate or Samsung 1TB HDDs for others. Variety of MOBOs (A320M, B450M, B550M) using Ryzen AM4 like 1600x, 2600, 2700X, 3600, 5600, or 5600G. Variety of memory but all have matched 16GB in two slots.
