I “think” what was going on was it was a barely working run that was flipping between 100/1000. Both ends at the time we’re unmanaged switches so I didn’t have a good way to check. Because of that I did swap the one to a little 8 port Cisco managed switch so I won’t fall into this trap again.
My winner of winners was flaky connection from our gateway router to my lab. It slowly degraded over time and got to the point I was lucky to get 10MB transfer speeds. Rebuilt my main server 3-4 times trying different things to fix it. Hooked up another server and it could talk to the first server fine. Thought it was a bad switch but everything tested fine until I found one link that wasn’t operating properly. Long story short some type of varment got ahold of my copper run in our craw space and damaged it enough to work but not well.