I was easily able to get the Ekos plate solving working with my guide camera (ASI120MC-S), but to date have failed to solve even a single image off the K-5iis. A post in the INDI forum gives a potential clue, which I thought I'd share with y'all (OK, mostly Greg :-) ):

In the options for the solver, there's a value called "downsampling". If you turn up the verbosity for the solver's logging, you can find out how many stars it's finding. According to the post, if you're seeing more than 1000 stars, the solver might bog down. In that case you can increase the "downsampling" value until the number of stars gets down around 200 or so, which should make the solver happy and carefree.

Can't wait to try it out. Of course by the time I've got clear skies, I'll have the new camera too (I HOPE!), but I might have the strength of will to dismount it in favor of the K-5iis for a few minutes...

Screenshot

I’ve decided to only connect my ASI120 mini to a RPI3/Stellarmate and my K-70 directly via USB to my laptop using Adams Tethering software, Andreas’ Pentax WiFicommander could be used as well, but power consumption could be an issue.
The ASI120 mini attached to a SW Evoguide 50ED will be used for platesolving and guiding (Stellarmate) seems to be working ok.

Good to know! I checked my settings and sure enough downsampling is turned on. Prolly why astrometry has been working well.