crozzbite SKILL: SaaS Logistics (Service Delivery)
[!IMPORTANT] Filosofía: "Uptime is Trust." Logistics in SaaS = Reliability + Support + Delivery Pipeline.
Core Concepts
| Concept | Definition | Standard |
|---|---|---|
| SLA | Service Level Agreement. Guarantee of uptime. | 99.9% (Triple Nine). |
| SLA Credits | Refund if uptime is breached. | 10-25% of monthly fee. |
| L1 Support | User help, resets, FAQs. | Chatbot / Junior Rep. |
| L3 Support | Code fix, DB outage. | DevOps / Senior Dev. |
| CI/CD | Pipeline to ship code. | Blue/Green or Canary. |
Critical Patterns
1. Uptime Tiers
- 99.9%: ~43 mins downtime/month. (Acceptable for most B2B).
- 99.99%: ~4 mins downtime/month. (Enterprise/Banking). Expensive to maintain.
- 99.999%: ~26 seconds/year. (Telco grade).
2. Support Response Time (Slo)
- Blocker (System Down): < 1 hour.
- High (Bug): < 4 hours.
- Normal (Question): < 24 hours.
3. Deployment Strategy
- Blue/Green: 2 identical envs. Switch traffic instantly. Zero downtime.
- Canary: Roll out to 5% of users. Monitor. Roll out to rest. (Risk mitigation).
- Feature Flags: Toggle features on/off live.
Anti-Patterns
- ❌ Deploying on Friday: Just don't.
- ❌ No Monitoring: Learning about downtime from angry users on Twitter.
- ❌ Single Point of Failure: One DB, one dev who knows the password.
- ❌ Over-promising: Guaranteeing 100% uptime (Impossible).
Instructions for Agent
- Define the Promise: "What is your SLA?" (Start with 99.5% or 99.9%).
- Structure Support: "Who wakes up at 3 AM if servers crash?" (PagerDuty).
- Deployment: Recommend Blue/Green or Feature Flags for B2B.