SLOconf, the annual conference laser focused on Service Level Objectives (SLOs), was held last week and featured four days of virtual sessions with a number of well-known speakers tackling major infrastructure and site reliability challenges for the DevOps community. With four tracks - SLOs for Everyone, SLO Stories, The Future of SLOs and SLO FUNdamentals - there was something for every skill and experience level. Nobl9, the company behind the conference, also introduced The State of Service Level Objectives survey, an update on the community-driven OpenSLO specification, and a new community initiative the Service Level Objective Development Lifecycle (SLODLC). 

VMblog caught up with Kit Merker, COO at Nobl9, to talk about how the industry is evolving and why the adoption of SLOs is so important to the foundation of Site Reliability.

VMblog: Tell me about SLOconf 2022.

Kit Merker:  This was the second year of SLOconf, and attendance was up more than 30 percent. Site reliability and where SLOs fit into a company's strategy are a hot topic because online presence and customer satisfaction are so tightly integrated. The conference tracks and our broad range of speakers ensured a really wide range of topics so that attendees could create a schedule that mapped to where their company is in the SLO lifecycle. We expected to always have a number of educational sessions, but we also explored more advanced topics as the use of SLOs increases and also expands within enterprises.

VMblog: That leads me to my next question. What were the findings of The State of Service Level Objectives survey?

Merker:  We weren't sure what we were going to find with the results, but thought it was the right time given the customer and industry conversations we have been having to get a sense of the maturity and use of SLOs from a range of enterprises. The State of Service Level Objectives survey received responses from more than 300 IT managers and executives and was conducted by Dimensional Research. What it found is that SLOs are growing in popularity with more than 8 out of 10 companies planning to increase their use. One prominent use case was companies using SLOs to provide visibility into their use of new technologies. For example, 87% stated using SLOs for microservices would increase their performance. We've made the full report available as a download.

VMblog: What industries use SLOs the most?

Merker:  SLOs have definitely moved the past few years from bleeding-edge where only the most technologically advanced with large teams were deploying them to now more leading-edge companies. While we expect the majority of SLOs to still be used purely for IT operations, the research showed that business teams (executives, manufacturing, R&D, marketing, finance, etc.) are increasingly using SLOs too. A large reason for the industry expansion we are beginning to see is that every company requires observability, but managing those tools can be resource-intensive. The survey showed that 35% of respondents use 11 or more observability and monitoring tools. As I like to say, observability without action is just storage, and SLOs can help optimize the use and results from those tools. This complexity is partially why we are continuing to contribute and promote new open source projects to provide clear specifications and best practices for how SLOs can be deployed and work with existing infrastructure.

VMblog: Which brings us to OpenSLO 1.0 and what it means for enterprises looking to deploy SLOs. Tell me a little bit about the concept and how the project works.

Merker:  We along with a number of core contributors introduced the OpenSLO specification last year at SLOconf 2021. This month, OpenSLO 1.0 was introduced. It's an open source project under the Apache 2 (APLv2) license, designed to be the industry-standard SLO specification, to make SLOs accessible to modern developer Git workflow, and providing a common interface for integration with the full ecosystem of cloud infrastructure, application monitoring and performance tooling. Developers around the world have contributed along with Nobl9 including GitLab, Lightstep, Red Hat, Sumo Logic and Tapico.io. Since this is a community-driven specification, we are always looking for new contributors. To learn more about contributing, developers can go to the OpenSLO GitHub page.

VMblog: A new open source project was also announced. What is SLODLC designed to do?

Merker:  Service Level Objective Development Lifecycle (SLODLC), is a repeatable methodology for creating metrics that matter for SLO users. It's a set of practices to aid in the process. We are collaborating with the SLODLC community on a handbook, templates, and examples to support the use of SLOs across industries. It's also under the Apache 2 (APLv2) license, and has strong support from leading companies with early contributors including Accenture, Etsy, Ford Motor Company, Furo, IAG, Oracle, OutSystems, Realogic Solutions, Trusted Shops, and Virtusa.

VMblog: This year has already seen a lot of ups and downs, how does the world market impact the adoption of SLOs?

Merker:  With the uncertainty in the world, budgets are coming under closer scrutiny and people are being asked to do more with less. SLOs tied to business outcomes help with this process as they tame inefficiencies by giving teams quantifiable metrics and shining a spotlight on areas that are not measuring up. With SLOs and error budgets in place, companies can focus on optimizing their existing tools and decreasing cloud and infrastructure costs.

##