PS Value
by Johannes Nicolai
1. Developer Productivity
1.1. CLI
1.2. Deploy Requests - non-blocking schema updates
1.2.1. no prod access to initiate schema change
1.2.2. review/preview
1.2.3. no replica lag
1.2.4. reversible
1.2.5. approval
1.2.6. (cancellable)
1.3. Web query console
1.4. Query performance view
1.5. Team/access right management
1.6. API
1.7. Database branching
1.8. Web based dashboards
1.9. Automated replica setup for reporting and data science use cases
1.10. UI / clean interface
1.11. (Materialized views)
1.12. Data branching
1.13. No pager duty - more time to develop
2. Expertise
2.1. Core maintainers, contributors and inventors of key DB and DevOps technology
2.1.1. Former DB SREs/SRE managers from
2.1.1.1. Google
2.1.1.2. GitHub
2.1.1.3. Booking
2.1.1.4. Twitter
2.1.1.5. Flipkart
2.1.1.6. Youtube
2.1.1.7. Percona
2.1.2. Vitess
2.1.3. orchestrator
2.1.4. pt-* tools
2.1.5. ghost
2.1.6. GitHub Actions
2.1.7. Kubernetes Orchestrator
2.2. Active support for the largest Vitess setups on the planet
2.2.1. Slack
2.2.2. GitHub
2.2.3. Twitter
2.2.4. Roblox
3. Security
3.1. Complete isolation of data bases
3.2. SOC II
3.3. RBAC on branch level
3.4. separate SSL client certs for every DB
3.5. GitHub Token Scanning partner
3.6. no manual passwords - only generated, unique access tokens per DB branch
3.7. only encrypted access (TLS)
4. Innovation
4.1. Only DB natively designed for Kubernetes
4.2. Inventors of DB branching
4.3. Inventors of ghost and orchestrate
4.4. Inventors of VReplication
4.5. Easy to understand, usage based pricing
4.6. Portals
4.7. Insights
4.8. (Boost)
5. Reliability / Maturity
5.1. MySQL used in production for more than 25 years
5.2. Vitess used in production for more than 10 years
5.3. PlanetScale is using PlanetScale to develop PlanetScale
5.4. Automated CI/Performance regression tests for Vitess
6. PaaS/SaaS/Serverless
6.1. we carry the pager
6.2. you pay by what you use, we determine the necessary infrastructure to optimally serve your requests
6.3. Web based dashboards
6.4. (Data Dog export)
6.5. you spend time on developing
6.6. we manage Vitess setup / maintenance
7. Database Protection
7.1. Thundering herd Protection / Query dedup
7.2. Hot row protection / Query sorting
7.3. Automated Result set/rows limitation (memory/CPU protection)
7.4. Unreasonable query guardrails
7.5. No DDL in prod branch possible
8. Scalability
8.1. Connection pooling
8.2. (Pod Autoscaling/Sizing)
8.3. (Sharding)
8.3.1. Automated replica setup
8.3.2. Sharding for data size
8.3.3. Sharding for read requests
8.3.4. Sharding for write requests
8.3.5. Sharding for GDPR (geo sharding)
8.3.6. Controlled sharding (chose dimension, adopt later)