2.0 KiB
2.0 KiB
| title | sidebar | showTitle | hideAnchor |
|---|---|---|---|
| Team Infrastructure & Deployments | Handbook | true | true |
People
- James Greenhill (Team lead, Data/Infra Engineer)
- Karl-Aksel Puulmann (Full Stack Engineer)
Mission
Make using and developing for PostHog as reliable as running water. Wherever you want it.
Goals
- We don't lose events
- Data is as up to date as possible
- Engineers always be able to ship and build
- Fail fast. Fix faster.
- Ship anywhere
- Stack scales with demand
- Support Small Teams (and contributors) in building and debugging Posthog
- Be frugal.
Responsibilities
Concrete things we take responsibility over:
- app.posthog.com and its infrastructure
- On Prem & Single Tenant deployments
- CI/CD - How we deploy
- Data infrastructure (Clickhouse, Kafka)
- Monitoring and Alerting stack
Customer
- Other Small Teams in making sure they have the tools (databases, queues, etc) and the ability to deploy effortlessly that they need to build
- End users (Both cloud and on-prem teams)
Output metrics
VPC
Retention
- Metric: Retention
- Objective: Better than cloud
Cloud
Data Loss
- Metric: Data loss %
- Objective: < 0.1%
Uptime
- Metric: Uptime
- Objective: > 99.99%
Speed
- Metric: Speed
- Objectives
- Event ingestion: TBD
- Query response: TBD
- Overall: We should anticipate increasing demand (either manually or automatically)
Cost
- Metric: Infra Costs
- Objective: Our costs should grow at a rate that is sublinear relative to scale
Dev Experience
Dev Experience NPS (Infra)
- Metric: Developer experience (relating to infra) (maybe NPS?)
- Objective: TBD (maybe NPS?)