Once the system "lost" a user_id var that says who started a batch, but it was a mandatory field that was filled at the beginning of the process.

I tried to find the reason, failed, called a senior for help, he also failed to find the reason.

After a couple of hours I looked to him, he looked back and I said,
"Let's call it a solar flare bit flip?"
"It's the only logical explanation"

So it became a bit flip issue to the PM, we had a good laugh, we send as "system instability" to the client.

The bug never happened again, no one ever found the true reason, maybe it was really a bit flip 🤔

  • 6
    I won’t even bother working on a bug unless it can be reliably reproduced. If something just happened once and never again… like you say, chock it up to a solar flare. I got better things to be wasting my time on. 🙂
Add Comment