Let’s say you are on a software development team without a dedicated operations team. All the developers are “responsible for production” and there is a weekly on-call rotation.

Is that being responsible for production?

For any non-trivial production system, there will be parts you are unfamiliar with. How can you be responsible for parts you do not understand?

OK fine. There are parts that you have personally built/improved and can take responsibility for. This is better. It narrows the scope of your responsibility, but it has only kicked the rock down the road. What does it mean to be responsible for this part of the production system?

Taking personal responsibility for part of a production system

The main focus of being responsible for a production system is uptime. The system is producing some business value, which is the purpose of keeping it running.

Production systems will edibility go down. Hardware will fail. Dependent services will go down. Bugs are written. One needs to prepare for these edibilities by:

• building/refactoring to simple resilient systems
• adding metrics/dashboard to facilitate finding the issue
• write runbook/scripts for oneself to prevent mistakes during an actual outage

Sounds like a bunch of work I don’t have time for

When you took personal responsibility for part of a production system, were you allocated any resources (I.E., time)? No? How can you be responsible if you don’t have time to do so?

By definition, if you are responsible for part of a production system, then your boss isn’t. If your boss is allocating resources, they won’t think of your responsibilities at all. They will only allocate resources to satisfy their own responsibilities.

To have personal responsibility for part of a production system, one must demand resources to maintain it. One must communicate to their boss the resources needed to satisfy one’s own responsibilities.

If you genuinely are responsible, your boss should not dictate how you do your work. If they push back on the resources you are demanding, ask them if they want to take over being responsible for the production system, as it seems they already know what it takes. Otherwise, one must assert their responsibility and the time required to do it. Responsibility without freedom is a trap.

Do you want both the responsibility and freedom to build a production system? You’re in luck, Battlefy is hiring.