At my current company (mid-sized start-up, ~30 engineers), we deploy once every week - our workflow looks like this:
feature branch -> Github PR -> Travis/Jenkins CI -> Code Review once tests pass -> merge branch to staging -> deploy to staging env every Monday morning -> QA -> deploy to production Thursday
There have been growing pains though, especially with QA. Reverts are often required, which means not just reverting one (squash-merged) commit but every concomitant commit as well. We're looking into continuous deployment/chatops/deploying feature branches a la GitHub, but another challenge is to speed up our tests. We have about 600 tests, that take well over 15 mins to complete (eek!). We're thinking about parallelizing the tests, in addition to mocking http/db calls out to shave off some minutes off these tests, but there's still a long way to go.
Would love to hear more about what your deploy pipeline looks like, your stack, the size of your company and especially about how you transitioned/plan to transition to continuous deployment.
feature branch -> Github PR -> Travis/Jenkins CI -> Code Review once tests pass -> merge branch to staging -> deploy to staging env every Monday morning -> QA -> deploy to production Thursday
There have been growing pains though, especially with QA. Reverts are often required, which means not just reverting one (squash-merged) commit but every concomitant commit as well. We're looking into continuous deployment/chatops/deploying feature branches a la GitHub, but another challenge is to speed up our tests. We have about 600 tests, that take well over 15 mins to complete (eek!). We're thinking about parallelizing the tests, in addition to mocking http/db calls out to shave off some minutes off these tests, but there's still a long way to go.
Would love to hear more about what your deploy pipeline looks like, your stack, the size of your company and especially about how you transitioned/plan to transition to continuous deployment.