Summary of "Reliability Engineering Mindset • Alex Ewerlöf & Charity Majors • GOTO 2025"

Summary — Reliability engineering mindset (Charity Majors & Alex Ewerlöf)

Core themes

SRE and organizational fit
- Google’s SRE materials are influential but describe a high‑resource, Google‑specific recipe.
- Most companies need “fit practices” rather than one-size-fits-all “best practices.”
- In many organizations, platform engineering often precedes full SRE adoption.
Ownership and feedback loops
- Modern reliable software depends on engineers owning code in production and keeping feedback loops short.
- Handovers and ambiguous ownership are frequent incident hotspots.
Trade-offs and negotiation
- Reliability costs real money and effort — higher availability (more nines) increases cost.
- SLOs/SLIs enable informed trade-offs and let teams push back on management with data rather than emotion.

Reliability involves trade-offs: achieving higher availability increases cost and operational complexity. SLOs let teams make those trade-offs explicit.

SLI / SLO guidance, pitfalls, and uses

Biggest wins
- The SLI/SLO concepts (introduced by Google) are powerful and broadly applicable.
- They normalize failure and enable the use of error budgets.
Common pitfalls
- Not measuring SLIs at all.
- Measuring the wrong thing (using the same metric for all services).
- Applying cookie‑cutter availability targets that can be weaponized or misapplied by management.
Practical advice
- Define a “service” as the business problem you solve (not just a microservice or database).
- Tailor SLOs to the service class and business impact — treat them as fit practices.
- Treat SLOs as negotiated contracts: surface the costs and trade-offs when higher reliability is requested.
- Use SLOs to depoliticize roadmap decisions and to prioritize work when error budgets are exhausted.

Tools, features, and workflows discussed

Alex’s open-source visual tool
- Front-end (SVG) visualization mapping provider → consumer dependencies.
- Helps list and prioritize failures by business impact and derive meaningful SLIs.
- Used at Volvo to scale SLO rollout.
Honeycomb features (advocated by Charity)
- SLOs as first-class citizens, computed from the same incoming event data used for investigation.
- Seamless exploration from SLO violations to raw events (bubble-up/diff workflows) so teams can pivot from metric to root cause without cross-tool data mismatches.

Observability trend

Observability 1.0
- Splintered stores for metrics, traces, and logs.
Observability 2.0
- A single‑source, columnar approach treating observability as a data problem.
- Enables richer queries across signals and reduces cross-tool mismatches.
- Next-gen tools are expected to treat SLOs as primary UI/entry points into systems.

Organization, scale, and rollout

Scaling SLO adoption
- Reaching hundreds of teams requires repeatable workshops, visual tools, and a common language (a goal of Alex’s book).
Org architecture and culture
- Organizational culture, team identity, and social dynamics strongly influence reliability practices.
- Alex’s book includes diagrams and models for org design and communication.
Role of engineering managers
- Managers should quantify and communicate costs, say “no” or propose alternatives, and move discussions from emotional asks to measurable trade-offs.

Concrete examples and anecdotes

Media company example
- Leadership requested 99.999% availability for a streaming app.
- Research showed users tolerated up to roughly 2 hours downtime (~99.7%), demonstrating that ultra-high nines were unnecessary and costly.
Career path and onboarding tip
- Alex uses TPOP (Tech, People, Operation, Product) as a personal onboarding framework.

Main speakers and sources

Charity Majors — co‑founder & CTO, Honeycomb
Alex Ewerlöf — senior software engineer at Volvo Cars; author of the SRE/SLO book discussed
Referenced works and vendors: Google SRE books, Alex Hidalgo (on SLOs), Honeycomb, Volvo Cars, and the Observability 2.0 movement

No further action.