Beyond the easy-to-reproduce bugs, there are some production issues (bugs, slowness, crashes) that are not easy to root cause. This post details how we setup observability within our production setup make it easier for us to root cause and quickly decipher the impact of such issues.

Cache refresh for popular API requests that are long-running and resource-heavy used to result in system unstability during the refresh-window. This post details this problem and the various solutions we evaluated. It also explains the solution we chose and the considerations while making our choice.

At one of my clients, our Sentry setup was reporting exceptions at the rate of 100+ exceptions per hour. This post details how we steered our Sentry setup to adequately organize this deluge of issues. And, how that, in turn, helped us effectively fix the issues.