A logging procedure for every layer
Production failures were being debugged on calls with no information. With the delivery head, I co-wrote a logging procedure for every layer of the exchange platform, and he took it to US leadership to decide on a rollout.
- Without it
- Support calls kept running without data, and failures stayed hard to locate.
- The call
- Fix the missing evidence first: agree what each layer must log before debating individual bugs.
- What I did
- Co-wrote the procedure with the delivery head over a few weeks. My review that quarter noted I was the first on the team to take ownership across layers.
- What changed
- The delivery head sent the draft to US leadership and asked a US architect to help finish it and decide on a rollout.
What people said
“We are getting into frustrating calls with no information and I think this document is something that we should start looking at this.”
“This is a draft but once implemented can give us greater insight into where things are failing that we have today.”