Hmm, confused on this. Is this just supposed to be more explicit about what is going on? I would expect this to use its own variable, not just copy the batch size configuration.

This class has batch logic to set the start + batchSize as if it was a limit and potentially do this for different batches across multiple threads. Most other places where restore indices is used doesn't have all this logic. For other places, it makes more sense to simply use the batchSize as a batch and run through all the aspects assuming the rest api doesn't time out. In the case of the CLL upgrade job there is no expected timeout so it will complete after fully iterating through all the aspects without having to use or implement this specific batching logic which is a bit more complex. Setting this one line allows this code to run as it was designed for now, but eventually I'd replace it with a simpler implementation. I don't have evidence one way or another whether the threading logic in this class is effective at the moment. Therefore preserving the way this class works for now.