UNPKG

voyageai-cli

Version:

CLI for Voyage AI embeddings, reranking, and MongoDB Atlas Vector Search

github.com/mrlynn/voyageai-cli

mrlynn/voyageai-cli

73 lines (51 loc) • 1.67 kB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
# Deployment Process Structured deployment process ensures smooth rollouts and minimizes risk. ## Deployment Pipeline ``` Code Commit → Build → Test → Sandbox → Staging → Canary → Production (git) (compile) (unit) (smoke) (full) (5%) (100%) ``` Each stage validates readiness before proceeding. ## Blue-Green Deployment Run two production environments: ``` Blue (Current) → Handles all traffic Green (New) → Idle, no traffic Deploy to green, test, then switch: Blue (Current) ← Traffic now here Green (New) ← Switches traffic from blue ``` Rollback is instant (switch traffic back to blue). ## Canary Releases Route small traffic percentage to new version: ``` Canary: 5% traffic to new version Monitor: Error rate, latency, crashes If OK: Increase to 10%, 25%, 50%, 100% If issue: Rollback to 0% ``` Catches issues before full rollout. ## Rollback Procedure If deployment fails: ``` 1. Acknowledge issue (page oncall) 2. Assess severity (errors, data loss?) 3. Execute rollback: kubectl rollout undo 4. Verify: Health checks pass, errors drop 5. Investigate: Why did deployment fail? 6. Fix: Address root cause before next attempt ``` Rollback should complete in <5 minutes. ## Deployment Checklist - [ ] Code review approved - [ ] Tests pass (unit, integration, e2e) - [ ] No secrets in code - [ ] Staging smoke tests pass - [ ] Performance benchmarks OK - [ ] Database migrations reversible - [ ] Runbooks updated - [ ] Team notified - [ ] Monitoring alerts configured ## See Also - [Environments](environments.md) - Environment configuration - [CI/CD](ci-cd.md) - Automated deployment pipeline