I personally think the change from master & slave was kind of silly, as far as I’m aware, it was a bunch of people with no background in CS who thought the application of the term to something that has neither race nor agency was an insult to black people.

But I digress. It led to better guidelines in the Linux kernel, which I think are useful. You should tailor the terms you’re using to the specifics of the task. If you have a master process that only has outward interfaces through the slave processes, you could use the term ‘director’ and ‘actor.’ if the master process is managing slave processes which compete over the same resources, you can use the terms ‘arbiter’ and ‘mutex holder.’ If the slaves do some independent processing the master does not need to know the details of, you can use the term ‘controller’ and ‘peripheral.’

Basically, use a term that is the most descriptive in the context of your program.

Edit: also, I don’t know why no one mentions this, but you can also use master/servant. Historically, there wasn’t a difference between servant and slave, but in modern days there is, so it’s technically different, technically the same.