This is exactly what happened to me because of a noisy power supply (and it was a 12V 5A one, so no low current issues, just noise). Looking at the logs it was clear that DEC and RA counts were off-limits (meaning off the expected number range) because of spurious bytes being received.
I would be very wary of just disabling the "abort tracking" option as a permanent fix. Communication problems also affect slews and you might end up with your scope trying to become one with the tripod.
Depending on which package you are pulling from, you may need to wait until the patch is integrated into a new version. Given that the PR was merged yesterday, the next new version you see when running your apt upgrade should have the change.
EDIT: I'm also in the process of fixing the indi-eqmod horizon management. Stay tuned.
You should be able to enter your horizon sets and have them properly asserted by the mount driver, promptly aborting tracking/slewing/goto-ing. That might help people with cramped observatory spaces
Note that the interpolation of altitude between horizon points is still linear, not spherical.
The warning? It's somehow innocuous I think: you did not define any horizon points, that's all. If the message bothers you, perhaps you could write the empty file from the INDI interface (and perhaps find another bug if the driver crashes while reading an no-data file?).