The 1 hour build time per env, resulting in a minimum 3 hours to go from dev to prod just in build time for a one line hotfix, is not acceptable. With episerver, we could build on stage, then push the package between stage and prod without rebuilding for prod, which would be pointless. Also builds took minutes vs hours. From a governance and security point of view, it would be far better to be able to push a tested package from say test to stage to prod, vs. building a potentially different one on each, with the associated unacceptable build times. We have no idea how we could optimise for build time, other than to remove functionality, which is not an option.
The simple outcome of this unworkable system is that if it takes 4+ hours to put a critical hotfix live, due to the slow build times, we will have to make changes directly to the production branch, and push it into production without any testing. We cannot afford to have our site down for 4 hours to make a one line change in a P1 production emergency. 4 hours = hundreds of thousands in lost revenue. With episerver we were able to write, build, test and deploy hotfixes within 1-2 hours, with full due diligence and due process. Luckily we only had to do this rarely (e.g. if a exploit was found, or a provider failed unpredictably).
If we add a more likely case, were we have to build on dev several times before we are ready to push a hotfix, then it could be 6-8 hours due to taking 1 hour each time. Building and testing locally is often ineffective as the local env is so different from the cloud env.
Add on to this the problem that the build on any env (including dev and prod) randomly fail due to Adobe issues (e.g. kubernetes issue), and we have to ask Adobe to fix, and adobe take anything from 1 week to 2 months to fix broken environments, and its a real concern.
