I was able to complete the guider calibration routine and start guiding, just as a test. I then moved to another part of the sky and thought I should run another guider calibration. The second and all other attempts failed. I could see the star move as RA drifting forward started and I could see it move back when RA drifting reverse started. However, the calibration routine never realized the star moved back to it's original position. It just kept drifting until I canceled the calibration.

I did have logging for Guiding enabled and when I looked at the log, I found a couple of things that seemed different about the calibration attempts that failed.

First, for the initial calibration run that worked, I would see messages like "Tracking Square Qret(297,207 32x32)" after each movement and the square seemed to move somewhat with the star. For the calibration runs that didn't work, I do not see any of these "Tracking Square" messages.

For the calibration runs that fail, the reported start position after each iteration is always the same.On my last run, using Auto Star, the initial position is reported as:But after each iteration, I get the 283.244, 226.719 position.
Now here is were it gets interesting. That position (283.244, 226.719) is the same value as the last successful autoguide operation before I stopped my test autoguide after the first successful calibration;It is acting like the current star position in the second and subsequent calibration attempts is not being updated and is left over from the last autoguide operation.

I am running Kstars 2.8.2 on a Mac with the indiserver running on a Raspberry Pi 2. I am using an SBIG ST2000 camera with the internal guide chip.

I am new to Kstars/Ekos/Indi, so I don't doubt this is user error of some sort, but any ideas?

Just in case, I have attached the guider log.

Thanks
Mark Ritchie