What version of OR-Tools and what language are you using?
Version: main, 9ed0074, 3194fd5
Language: Java

Which solver are you using (e.g. CP-SAT, Routing Solver, GLOP, BOP, Gurobi)
CP-SAT

What operating system (Linux, Windows, ...) and version?
Linux

What did you do?
Steps to reproduce the behavior:

1. Solve the attached model using 9.9.4132 (9ed0074) (it takes ~20 seconds on an i9-10900K)
2. Solve the attached model using 9.9.4133 (3194fd5) (it takes ~40 seconds on the same hardware)

What did you expect to see

log_9ed0074090aacba704afd420234a6083e2644b5b.txt

What did you see instead?

log_3194fd5bb2017957dd3460719ba08affaa803511.txt

The model I used for comparison:
model.zip

Some differences I noticed in the logs:

• # vars in presolve went down (different propagation changing probing order?)
• # propagations under "Search stats" went up, but the summary stayed the same
• approx gap integral went up
• all SAT stats increased significantly