GitHub recently launched an engineering blog, and while they haven’t had a ton of posts yet, some of the first blog entries are really neat. In a May 19, 2015 article, Nathan Witmer looked at MySQL’s performance schema. Witmer’s findings were incredibly revelatory.
The performance schema includes a neat feature that tracks unique queries and how often they are executed. According to the Witmer’s analysis, the folks over at GitHub discovered that one update was handling more than 25% of the updates on one of their heftier tables. Apparently, whenever a health status check executed on a repository, the column in question updated. The GitHub team then analyzed how often the needed to be updated. The result: less than 5% of the time. A quick code tweak, and the excessive updating was therefore remedied.
Check out the full post on the GitHub blog. In their most recent post, GitHub’s Emily Nakashima offered a fascinating GitHub’s browser monitoring. The best part about the blog: it presents real-world use cases. Such write-ups incredibly useful, especially since the tips, tricks, and tutorials are coming from the well-respected GitHub.
Pop on over to the blog, as it’s certainly an awesome resource, and one we’re pretty jazzed up about over at DZone.