As much as we’d love to blame yesterday’s outage on scaling issues, hardware, exponential growth, rogue bots, aberrant behavior, or even our brief stint on Japanese television last night, we can’t. Having achieved a strong position with regard to scaling Twitter, we felt comfortable enough to begin optimizing lots of smaller parts of our application for maximum efficiency. We were so focused on those smaller details that we lost track of the bigger picture and the site was unresponsive for lots of folks throughout the day.Read more…