Would you be able to split your content out across multiple indexers? For example, if you are storing documents for a profession sports league, rather than lump all of the documents into a single indexer for the entire league, split the documents up so that
each team has their own indexer. That way, you lower your risk of running the same indexer. More info
here

You can speed up indexing by parallelizing it: split blobs in your container into several folders, and create several datasource / indexer pairs all writing into the same target search index. If your search service has N search units, N indexer can be running
concurrently, resulting in a significant speed-up of indexing.
Source