How are we transforming the org to achieve 99.99% uptime at DigiCert
summary
DigiCert pursues a concrete target: raise uptime from ~99.5% (~3.65 hours monthly downtime) to 99.99% (under five minutes). The team adds endpoint health checks wired into New Relic and Splunk to shrink detection time, applies stability patterns (timeouts, bulkheads, circuit breakers using Resilience4j) to stop cascading failures, and builds blue‑green Kubernetes clusters so DB‑compatible schema changes and rapid rollbacks cut mean outage minutes to roughly one third year‑over‑year.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 1:04DigiCert overview
- 3:09High availability project
- 4:35Lesson one: Don't go too big
- 7:46Measuring uptime
- 10:13Lesson two: Domino effect
- 14:28Lesson three: Bad deployments
- 18:47Downtime reduction
- 19:48Three key lessons
- 20:10Q&A
- 21:58Q&A: Downtime definition
- 23:27Q&A: Service dependencies
- 24:55Q&A: 99.99% uptime
- 26:20Q&A: Blue-green deployments
- 28:05Q&A: Book recommendations
related talks
30mZero Down Time Operations: How to Get Atleast 3 9's of Uptime
46mMeeting Impossible SLAs: How we made our uptime 99.999%
31mHow We Went From Two Major Outages to 99.98% Reliability in 6 Months
56mGetting all the 99.99(9) you always wanted
56mHow to lay the groundwork in advance, or the path to 99.99% project uptime
13mAutomating Certificate Management with
20mEvangelizing the SRE mindset
26mUptime, Mean-Time, and Ahead of Your Time