When I'm learning data structure, I find that people only care about how to do read operations on graph data structure, such as breadth first search, depth first search, minimum spanning tree, single source shortest path (SSSP) problem. However, there's limited resource about how to update a graph, such as put, get, delete, which has always been studied in data structures like hash table, or tree structure.
I think if one graph is large enough, say 100GB, these updating operations could bottleneck the performance. You must encure the consistency in concurrency, and you may want to avoid random read operations and etc.
Therefore, why does nobody care about updating a graph? Is it trivial? Or is it too hard to optimize the performance? Thanks!