So, I should just throw Redis at it? Then it will be ROFLScale and I can sleep at night? I'm trying not to be a complete prick about this, but how could you even begin to make the assumption that our database is our bottleneck without knowing a single thing about our application. How do you even know we're not using Redis already?
Our bottleneck is the computational complexity of our datasets. The nature of reverse auctions is that the dataset changes frequently. The fact that every bidder is ranked on a per line-item basis means there are cache interdependencies, which cause tidal waves of cache expiration when bids are placed. Adding difficulty, there is a lot of pressure to deliver a subset of the data in the same request/response cycle, so that bidders aren't waiting in limbo for their rank. This helps maintain bidder momentum.
We don't currently have any scaling issues that we haven't solved. That's not the point. The point is that suggesting blanked solutions to problems you don't understand is major Dunning–Kruger territory.
Our bottleneck is the computational complexity of our datasets. The nature of reverse auctions is that the dataset changes frequently. The fact that every bidder is ranked on a per line-item basis means there are cache interdependencies, which cause tidal waves of cache expiration when bids are placed. Adding difficulty, there is a lot of pressure to deliver a subset of the data in the same request/response cycle, so that bidders aren't waiting in limbo for their rank. This helps maintain bidder momentum.
We don't currently have any scaling issues that we haven't solved. That's not the point. The point is that suggesting blanked solutions to problems you don't understand is major Dunning–Kruger territory.