Support Your Application Engineers

Hey folks,

I have a Platform Engineering story about the recent experience of a close climbing buddy who will remain anonymous. I shared this story on my LinkedIn, but wanted to share it here too.

Let’s call my friend “Bob”.

Bob is an application architect and engineer at a recently IPO'd company. He was telling me about a frustrating recent interaction that he had with his infra team.

Bob needed an S3 bucket to store some files for his application. This story is especially comical since S3 bucket provisioning is my go-to "self-service" example—the costs are low, the operating complexity minimal.

Bob was told that he needed to submit a PR against the org's terraform monorepo. The infra team didn't have time to help him with his request. Fair enough, we’re all busy.

Bob took the time to figure out the terraform codebase without any help from the infra team or internal documentation. He submitted the requested PR and then pinged the infra team for help, review and ultimately to get his S3 bucket provisioned.

Can you guess what happened next?

The infra team:

• rejected his PR

• didn't give him actionable feedback

• and ultimately told him that he should basically go pound sand

He was frustrated as hell.

So was I when he was telling me.

Bob is one of the few engineers genuinely open to learning IaC so he could move his product forward instead of waiting in line for a ticket to be completed.

He put in a solid effort to learn TF and try to ship the infra he needed. And yet Bob wasn't supported. At. All.

Ultimately it was a horrible experience for him. The sad thing is that all Bob needed was a bit of help and a few pointers in the right direction to make sure his PR was correct. Yet his infra team completely dropped the ball.

Moral of the story: support your damn application engineers! Do so by enabling them to make changes and ensuring they have a clear, documented process to follow.

Don't complain about your backlog being too long if you're gate keeping your organization and stopping smart engineers from doing simple self-service infrastructure operations. When that happens, you're the cause of failure, not the backlog or the application engineers.

May your infra team always be helpful, Matt @ Masterpoint

PS If you’ve found this newsletter helpful, please forward it to a friend. I’d truly appreciate it! And if you ever want to talk shop about some tricky IaC issue that you or your org are running into: Grab some time on my calendar here.