Everything is on fire: Production outages, database crashes, poor performance.

Unplanned downtime is more frequent, possibly affecting your SLAs. And when it’s not, your customers are showing up at your support desk, frustrated by high or unpredictable latency. That stresses out management, which stresses out your development team.

And then one day you realize your team is doing more firefighting than actual constructive development. This is one of the clearest indicators that your development is off-track.

“Let’s throw it all out and start over,” someone says. Maybe. Greenfield rewrites are tempting, but don’t assume that’s the easiest way out. There are many techniques for upgrading by layers, for updating component by component, and for isolating certain sections of the software that are the most difficult to upgrade.

But the first priority is to stop the building from burning. Your current software version needs to be stabilized—fixing erratic performance and stopping it from chewing through costly infrastructure resources.

I’ll define the project first by answers to these questions:

• What’s breaking right now?

• How bad is it?

• Who’s affected?

• What have you tried so far?

• What layers or components of the software have the most performance bottlenecks, require the most firefighting or drain the most resources?

I’m very familiar with how urgent matters can get in a short amount of time—use my contact form and I’ll quickly make time to discuss short-term triage matters.