We’ve all been in the situation where a client project comes up, needs to happen quickly, efficiently, and with the lowest overhead possible since a testing site/application might never come to fruition as decisions will likely change on-the-fly. A run-of-the-mill L.A.M.P. config / setup might do the trick, but if you’re not optimized it’s going to show.
What if we can do a L.A.M.P. configuration without actually using it? Is it possible to offset the stack outside of the normal “stack” methodology?
Testing environments aren’t always the best and certainly end their share of not-so-optimized performance and reliability. So how does one find the proper balance between low-cost and more than decent performance for the client to review?
We came across a pretty pretty interesting article from Startupgrind that looks like it leverages all the right theory and application. Take a look.