For read heavy workloads, https://docs.rs/papaya/latest/papaya/ should be pretty good. It's a proper concurrent hashmap, with incremental resizing. It even has some support for rcu operations per field