Sometimes, you need to deploy a change you're a little worried about. Sure, it passed the automated testing, and it worked fine in preproduction, but once you put it out there in the real world, it gets exposed to users who do weird things. Wrong things. Stupid things. But they're still your users, and you care about them, so you want to make sure their crazy, unpredictable inputs are handled…
Keep Reading →
Picture the scene: something broke in production. If you're doing it right, it was detected by your monitoring tools. If you're doing it wrong, your users helpfully informed you: by email, on the phone, through Twitter. You desperately scramble around for a solution. You roll back a release, modify some configuration, reboot a machine. Something worked; production is back up. The phones stop…
Keep Reading →