test-a-zero-downtime-deploy
Test a zero-downtime deploy across mixed versions, readiness, drain, schema compatibility, sessions, jobs, traffic, rollback, and user outcomes. Use when a release promises uninterrupted service.
来源信息
- 仓库
- protosphinx/sphinxstack
- 最近来源活动
- 2026年7月29日 13:55
- 检测到的 SKILL.md 语言
- 英语
- 星标
- 3
- Forks
- 0
安装方式
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
检查来源文件
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
文件资源管理器
2 个文件正在显示 SKILL.md
SKILL.md
来源说明 · 只读预览- name
- test-a-zero-downtime-deploy
- category
- code
- description
- Test a zero-downtime deploy across mixed versions, readiness, drain, schema compatibility, sessions, jobs, traffic, rollback, and user outcomes. Use when a release promises uninterrupted service.
# test-a-zero-downtime-deploy
## When to use
- Use in production-shaped staging and a bounded canary.
- Do not equate successful process replacement with uninterrupted user journeys.
## Procedure
1. Define critical journeys, availability, error, latency, and data-integrity gates.
2. Verify old-new API, event, schema, cache, session, and job compatibility.
3. Generate steady traffic through rollout, drain, startup, and termination.
4. Track requests, streams, transactions, jobs, and side effects across instances.
5. Test node loss, slow startup, failed readiness, partial rollout, and rollback.
6. Reconcile user outcomes and data before declaring success.
## Failure plan
- Stop and roll back when mixed-version or user-journey evidence fails.
## Worked example
HTTP stays green while long-lived uploads reset during drain, disproving zero downtime.
## Done
- A deploy test report records journeys, versions, traffic, readiness, drain, schema, sessions, jobs, rollback, and outcomes
- Mixed-version, stream, transaction, side-effect, node, readiness, rollback, and reconciliation tests verify continuity
在 GitHub 查看