Everything else being equal, I might consider Python to leverage both its multi-threading capabilities (for the scraping) and its comprehensive database tools. But really pretty much any of the popular server-side languages could do it, with the database design/implementation perhaps being more important in terms of performance and scalability.