Bookmarks (10) clear filters

Scaling Datastores at Slack with Vitess Slack serves 2.3 million QPS at peak. 2M of those queries are reads and 300K are writes. Median query latency is 2 ms, and p99 query latency is 11 ms. This shareds detail how they evaluate Vitess and migrate to it at their scale. They are 99% on track of migrating to Vitess at this point

Building the Next Evolution of Cloud Networks at Slack Network is the backbone of almost any cloud application. When companies scale, we add more network, qa, staging, prod, support, rds, elasticache. Eventually we need to allow those network to communicate with each others. How to manage that complexity? Adding a new network and allow traffic flow require you to add a new route table....

Development Environments at Slack A look at how Slack setup development environment. They actually use cloud for dev environment instead of running them on developer laptop because the Slack app is too complicated and they try to mimick production as much as possible.

Deploys at Slack This post will outline their process and a few of the major projects that got them to where they are.

How Slack Built Shared Channels “A shared channel is one that connects two separate organizations. Let’s look into architecture of shared channels: sharding, storage, privacy.”

Chrome Tracing for Fun and Profit Love or hate it. You have to write JavaScript and use devtool to debug it. Slack write up on using Chrome Tracing: a tool that’s built into Chrome (and Electron) that can collect a huge variety of detailed performance data.

Evolving API Pagination at Slack Pagination is a fundamental features of almost all web application. Paginate without duplicating item or missing new item isn’t trivial.

Keep webpack Fast: A Field Guide for Better Build Performance webpack is a brilliant tool for bundling frontend assets. But when project becomes large, it becomes slow. Slack teams improve this with measuring to identify bottle neck, parallelize, reduce the workload.