I would strongly disagree with any suggestion that these predictably repeatable changes in clock rates are merely a product of clock "malfunction," though.
Old cars lack of sensors and a computer. You were driving with an almost flat tire, with very old and burnt motor oil, with a blown brake light, and you never noticed it... sometimes until it was too late.
With new technology, you can "perceive" the several troubles and failures in your car because the new technology helps you to detect it.
The same happens when the clocks became more accurate. The several "failures in the machine" are detected now. something that was not possible with other mechanical and digital clocks -which also malfunctioned- because they didn't show data at the nano level.
But, a sand clock and a water clock in outer space will also show you that it is not time
but the clock itself suffering malfunction when going to outer space, when they are exposed to motion, and so forth.
Look, when L Essen invented the atomic clock, the relativists never told him that they have already predicted the slowing of time and that his invention will prove it.
On the contrary, the relativists asked him his permission to make some experiments with the atomic clock.
In those years, nobody predicted the amount of difference between the data given by a clock in outer space and a similar clock standing on ground zero.
The difference of data was known after the clock was put in orbit.
This was a "trial and error" scenario.
After it was noticed that it was a difference with the received data from clocks in outer space, that difference was calculated, and another receiver was created to "update" the received data and make it to agree with the data given by the atomic clocks on ground zero.
L. Essen, mocked on relativists, he considered them like idiots making fool to themselves and to generations of scientists.
How the changes in the malfunction happen to be "the same"?
Simple, just buy cheap digital watches, same brand, new batteries, and put a few of them in the freezer compartment of your refrigerator. Keep one of the cheap watches hanging in the wall of the living room. After one day, the watches inside the freezer compartment will show "slow data" of 5-6 seconds delay. You can check the watches inside the refrigerator everyday and this delay of 5-6 seconds will be regular.
Amazingly, you take the watches out of the freezer compartment, and they might return back to give their data as they did before the experiment. In other words, they will return back to their normal functional work.
This is a proven experiment, made by an elementary school student.
You won't beat what this student found out. I can guarantee it.