The Temporary Pipe That Stayed

A “temporary” sync script ran for 18 months. It was quietly eating 20% of the compute bill.

I found it during a platform review at a scaleup: a script on a forgotten VM, on a cron, copying full tables between systems every 2 hours. Written as a stopgap for a launch. The engineer who wrote it had left a year earlier.

Nothing about it was broken. That’s why it survived: it ran, nobody owned it, and the cost hid inside a warehouse bill that was growing anyway.

The fix was unglamorous. Incremental loads instead of full copies, daily batch instead of every 2 hours (nobody could name a consumer that needed fresher), an owner in the on-call rotation. Compute dropped roughly 20%. The review paid for itself several times over before it ended.

The pattern I now assume by default: every stack has at least one of these. Temporary code outlives its author unless it gets an expiry date or an owner. Usually both are missing.

What’s the oldest “temporary” solution still running in your production environment?