A lot of outages off late seem to be related to automated config management. Companies seem to place a lot of trust is configs being pushed automatically witho

A lot of outages off late seem to be related to automated config management. Companies seem to place a lot of trust is configs being pushed automatically without human review into running systems. Considering how important these configs are, shouldn't they perhaps first be deployed to a staging/isolated network for a monitoring window before pushing to production systems? Not trying to pontificate here, these systems are more complicated than anything I have maintained. Just trying to think of best practices perhaps everyone can adopt.