Guido Iaquinti

From the very beginning of Slack, MySQL was used as the storage engine for all our data. Slack-operated MySQL servers in an active-active configuration. This is the story of how we changed our data storage architecture.

KubeCon 2020, virtual conference.

Have you ever considered migrating a database system at scale with no downtime? Many of us that have tried often find it an insurmountable challenge for developers and database engineers. In this talk, Rafael and Guido will discuss how they designed and built a migration framework and then executed it to move petabytes of data to Vitess with zero downtime.

KubeCon 2019, San Diego, CA, USA.

Monitoring and observability are essential concepts, especially in complex and distributed systems. Redundancy and defensive programming are also necessary, but sometimes they are not enough. Designing systems to minimize the blast radius when the unexpected happens is often the key.

In this talk, Rafael and I will share an overview of how Slack was designed, built, scaled, and then iterated to improve its distributed database service based on top of Vitess, now a CNCF project. The databases team at Slack rose a Vitess cluster from 0 to spikes of 2.7 million queries per second. This journey has taught us how to operate a database cluster with more than 2.000 nodes, and we expect growth to more than 3,500 in the next 12 months.

PerconaLive Europe 2019, Amsterdam, The Netherlands.

Traditionally, database reliability has focused heavily on the stability of a single server or a small number of servers. At Slack, however, we've built a database architecture that instead focuses on an approach of strength in numbers. This talk will review the architecture we've made, specifically focusing on our Vitess infrastructure.

PerconaLive 2018, Santa Clara, USA.

Slack is a messaging platform for teams that brings all types of communication together, creating a single unified archive accessible through powerful search.

MySQL is the primary storage for all our customer data, and we currently execute billions of transactions per hour. As more users join the service and Slack becomes a more critical part of their workflow, the system becomes more complicated to manage. What started as a simple MySQL database was only the starting point for redesigning our entire database infrastructure.

This talk analyzes how our operations team took Vitess, a bleeding edge, poorly-documented open-source software developed by Google, and then hardened, tested, and shaped it for our infrastructure and hosted all our mission-critical data. This presentation will consider the technical challenges we faced to deploy this project successfully, the key decisions we took, what went well, what didn't, and the course correction we made along the way.

Attendees can expect to hear details about how we took some whiteboard conversations and turned them into battle-tested, production-caliber systems.

DevOpsDays 2017, Tel Aviv, Israel.

The journey of a small operations team in a fast-growing environment is always challenging. This talk describes Slack's solutions to many of our technical and cultural challenges as we scale our technical operations team worldwide.

We analyzed key decisions, then dug into what went well and what lessons we learned along the way. Attendees can expect to hear details about issues such as handoff between time zones, partnering with software engineers for significant projects, and dealing with long-running incidents.