I cannot see your output file, but I am pretty sure that the program is running into an error when starting the optimization i.e. when it is trying to generate the guess hessian.
You are using Cartesian coordinates for optimization, and GAMESS by default will attempt to eliminate the rotational and translational modes from the guess hessian, before the optimization starts. The algorithm that does this sometimes runs into problems. For non-linear molecules, there should be 6 rot./trans. modes, and for linear molecules 5 rot./trans. modes. If the algorithm cannot find 5 or 6 modes which match its criteria, then it throws this error and GAMESS stops.
There are two ways to solve this:
You can use internal coordinates to optimize. In this case, to do this, you need to add
$CONTRL and also add
$ZMAT DLC=.t. AUTO=.t. $END which will instruct the program to use internal coordinates, and generate delocalized internal coordinates, respectively. The guess hessian will be generated in internal coordinates from the start, removing the need to eliminate rot./trans. modes.
You can force the optimization to proceed without eliminating the rot./trans. modes. Add
$STATPT for this. I have done this a few times, and haven't seen any problems so far, but you should be careful.
I would recommend the first choice in general, because optimizations in internal coordinates are often faster. Of course, in this case your molecule is so small that you probably won't notice the difference.