Production Readiness Reviews
যে PRR চেকলিস্ট প্রতিরোধযোগ্য launch-into-fire-এর 80% ঠেকিয়ে দেয়, একটি বাস্তব launch gate Terraform module সহ।
বাস্তব জগতের উদাহরণ
occupancy-র আগে একটা building inspection — মানুষজন ঢোকার আগে কোনো নিরপেক্ষ কেউ যাচাই করে দেখে যে structure-টা নিরাপদ।
গল্পে বুঝি
ফাতিমা আল-ফিহরি শহরের বিল্ডিং সেফটি ইন্সপেক্টর। শহরের প্রান্তে একটা নতুন কারখানা সবে তৈরি হয়েছে, মালিক আজই শ্রমিকদের ঢুকিয়ে উৎপাদন শুরু করতে চান। কিন্তু ফাতিমার occupancy certificate সই না হওয়া পর্যন্ত কারখানার গেটে তালা। তিনি হাতে একটা ছাপানো চেকলিস্ট নিয়ে ভেতরে ঢোকেন, আর এক এক করে যাচাই করেন — fire alarm আর smoke detector কি ওয়্যার করা এবং টেস্ট করা আছে? জরুরি বেরোনোর পথগুলো কি খোলা এবং চিহ্নিত? দেয়ালে টাঙানো লিখিত emergency drill-book কি জায়গামতো আছে? প্রতিটা শিফটের জন্য কি প্রশিক্ষিত লোক roster-এ বসানো? আর wiring কি পুরো লোড সামলাতে পারবে, নাকি সব মেশিন একসাথে চললে গলে যাবে?
আজ smoke detector লাগানো আছে ঠিকই, কিন্তু কোনোটাই টেস্ট করা হয়নি, আর দ্বিতীয় শিফটের জন্য কোনো প্রশিক্ষিত লোক roster-এ নেই। ফাতিমা certificate সই করেন না — “এই দুটো ঠিক করুন, তারপর আবার আসব।” মালিক বিরক্ত হন, একদিন দেরি হয়। কিন্তু ফাতিমা জানেন, চালু কারখানায় আগুন লাগার পর detector কাজ করছে না আবিষ্কার করার খরচ আজকের এই দেরির চেয়ে হাজার গুণ বেশি।
এই গল্পটাই আসলে Production Readiness Review। ফাতিমার occupancy sign-off = service launch করার আগের PRR gate। টেস্ট করা fire alarm আর smoke detector = monitoring ও alerting; দেয়ালের drill-book = runbook; প্রতি শিফটে প্রশিক্ষিত roster = on-call coverage; পুরো লোড সামলানো wiring = capacity (আর তার সাথে SLO দিয়ে মাপা নির্ভরযোগ্যতা); চিহ্নিত জরুরি বেরোনোর পথ = graceful degradation/failover। যেকোনো একটা item fail করলে launch block। বাস্তবে Google-এর SRE team ঠিক এভাবেই নতুন service-এর pager হাতে নেওয়ার আগে একটা structured PRR চালায় — একটা মিসিং dashboard launch-এর সময় ঠিক করা, 2am-এর incident-এ ঠিক করার চেয়ে ঢের সস্তা।
PRR কেন থাকে
Production Readiness Review (PRR) হলো একটা structured checkpoint, যেটা SRE কোনো service-এর pager গ্রহণ করার আগে হয়। এটা একটা কারণেই থাকে: launch-এর সময় একটা missing dashboard ঠিক করার খরচ, 2am-এর incident-এর সময় সেটা ঠিক করার খরচের চেয়ে 10x কম।
PRR হলো SRE যা করে তার মধ্যে সবচেয়ে leveraged জিনিস। এক ঘণ্টার review কয়েক সপ্তাহের pager-যন্ত্রণা ঠেকিয়ে দেয়।
launch criteria checklist
একটা PRR হলো yes/no checklist। যেকোনো item-এ অস্পষ্টতা মানে launch block, যতক্ষণ না সেটা সমাধান হচ্ছে।
# PRR: [Service Name] — [Owner Team] — [Launch Date]
## 1. Architecture
- [ ] Architecture diagram exists and is current
- [ ] All dependencies documented (services, databases, external APIs)
- [ ] Failure mode for each dependency documented
- [ ] Single points of failure identified and accepted (or eliminated)
## 2. Reliability
- [ ] SLI defined and measurable in production
- [ ] SLO target agreed by product + SRE, written down
- [ ] Error budget policy signed off
- [ ] Capacity plan completed (Little's Law math + headroom)
- [ ] Load test passed at expected launch traffic + 2x
## 3. Observability
- [ ] RED metrics for every endpoint
- [ ] USE metrics for every infrastructure component
- [ ] Structured logging with correlation IDs
- [ ] Distributed tracing for at least the critical path
- [ ] Dashboard exists, linked from the runbook
- [ ] Cardinality budget reviewed (no unbounded labels)
## 4. Alerting
- [ ] Symptom-based alerts on SLO burn rate
- [ ] Every alert has a linked runbook
- [ ] Alert threshold tested in staging (false-positive rate under 5%)
- [ ] Escalation path defined in PagerDuty
## 5. Deploy + Rollback
- [ ] CI runs unit + integration + smoke tests
- [ ] Deploy is automated (no manual steps)
- [ ] Canary stage with at least 5% traffic for 30 min
- [ ] Rollback is one command and tested
- [ ] Database migrations are forward + backward compatible
## 6. Security
- [ ] Secrets in vault, not in env or repo
- [ ] mTLS or equivalent for service-to-service auth
- [ ] AuthZ rules reviewed by security team
- [ ] Dependency vulnerability scan clean
- [ ] PII handling reviewed (if applicable)
## 7. Operations
- [ ] Runbook exists with at least: alert→diagnosis→mitigation
- [ ] On-call rotation populated and acknowledges service ownership
- [ ] Backup + restore tested in the last 90 days (if stateful)
- [ ] DR runbook exists (RTO + RPO documented and tested)
- [ ] Cost forecasted for steady-state and 10x scale
## 8. Launch
- [ ] Feature flag exists for emergency disable
- [ ] Status page entry created
- [ ] Customer support team trained on common questions
- [ ] Post-launch monitoring shift assigned (extra eyes for 48h) এটা একটা বিস্তৃত list। এটাকে নিজের প্রয়োজন অনুযায়ী সাজিয়ে নিন — বেশিরভাগ service-এর প্রতিটা item লাগে না, কিন্তু team-কে সচেতনভাবে কোনো item skip করতে হবে, ভুলে গিয়ে বাদ দিলে চলবে না।
একটা বাস্তব service-এর জন্য একটা worked PRR
চলুন একটা payment-service launch-এর জন্য একটা PRR চালাই। বাস্তবে যেমনটা হতো ঠিক তেমন কথোপকথন:
SRE: "Walk me through the deploy."
DEV: "We git push to main, GitHub Actions builds, deploys to staging,
runs smoke tests, then rolls to production."
SRE: "What if smoke tests pass and prod is broken?"
DEV: "We... have not actually tested rollback. We'd revert the commit
and redeploy."
SRE: "How long does that take?"
DEV: "Maybe 8-10 minutes for the full pipeline."
SRE: "BLOCKED. Rollback must be one command and complete in under
2 minutes. Add `kubectl rollout undo` to your runbook,
test it Monday, then we re-review." সেই কথোপকথন, পুরো checklist জুড়ে বারবার করলে সেটাই হলো PRR। প্রতিটা block আসলে একটা বাস্তব জিনিস, যেটা team শেষ পর্যন্ত একটা বাস্তব outage-এ অনেক বেশি খরচে আবিষ্কার করত।
checklist-কে policy হিসেবে codify করা
যে checklist মানুষ চালায়, সে checklist মানুষ skip করে। এটাকে deploy tooling-এর ভেতরে bake করে দিন।
# terraform/modules/service/main.tf
# A real internal module that enforces PRR baseline at infrastructure level
variable "service_name" { type = string }
variable "owner_team" { type = string }
variable "slo_target" { type = number } # e.g., 0.999
variable "runbook_url" { type = string }
# Enforce SLO is realistic
locals {
validate_slo = var.slo_target > 0.99 && var.slo_target < 0.99999 ? null : (
file("ERROR: slo_target must be between 0.99 and 0.99999")
)
}
# Mandatory: PagerDuty service must exist
resource "pagerduty_service" "this" {
name = var.service_name
escalation_policy = data.pagerduty_escalation_policy.team.id
alert_creation = "create_alerts_and_incidents"
}
# Mandatory: SLO recording rules
resource "kubernetes_manifest" "slo_rules" {
manifest = yamldecode(templatefile("${path.module}/slo-rules.yaml.tpl", {
service_name = var.service_name
slo_target = var.slo_target
}))
}
# Mandatory: burn-rate alerts wired to PagerDuty
resource "kubernetes_manifest" "burn_alerts" {
manifest = yamldecode(templatefile("${path.module}/burn-alerts.yaml.tpl", {
service_name = var.service_name
slo_target = var.slo_target
pagerduty_service = pagerduty_service.this.id
runbook_url = var.runbook_url
}))
}
# Mandatory: dashboard provisioned
resource "grafana_dashboard" "service" {
config_json = templatefile("${path.module}/dashboard.json.tpl", {
service_name = var.service_name
})
folder = data.grafana_folder.team.id
}
# Mandatory: runbook URL must respond 200
data "http" "runbook_check" {
url = var.runbook_url
lifecycle {
postcondition {
condition = self.status_code == 200
error_message = "runbook_url ${var.runbook_url} returned ${self.status_code}"
}
}
} নতুন একটা service launch করতে হলে, team-কে এই module invoke করতেই হবে। একটা SLO, একটা alert, একটা dashboard, আর একটা reachable runbook ছাড়া তারা production-এ কোনো service তৈরি করতে পারবে না। PRR “checklist” থেকে “compile-time error”-এ চলে এসেছে।
সঠিক পথটাকেই সহজ পথ বানান। যদি আপনার launch tooling PRR-এর 80% automatically enforce করে, তাহলে human review সেই 20%-এ মনোযোগ দিতে পারে যেখানে বিচারবুদ্ধি লাগে (capacity, security, customer impact)।
runbook (সবচেয়ে বেশি skip হওয়া item)
Runbook হলো সেই artifact যেটা বেশিরভাগ team সবচেয়ে খারাপভাবে বানায়। একটা বাস্তব runbook একটা structured shape মেনে চলে:
# Runbook: payment-service / PaymentLatencyHigh
## Alert
PaymentLatencyHigh — p99 latency > 500ms for 5 minutes.
## Severity
SEV2 if duration < 30 min. SEV1 if revenue impact > $1k/min.
## Owner
Team: payments-platform
Slack: #payments-oncall
PagerDuty: payments-primary
## What this means for users
Checkout still works but feels slow. Cart abandonment may rise.
## Diagnostic steps
1. **Check the dashboard:** [direct deep link](https://grafana.example.com/d/abc/payment-service?from=now-1h)
2. **Identify which endpoint:** Look at the "p99 by endpoint" panel.
3. **Check recent deploys:** `kubectl rollout history deployment/payment-service`
4. **Check upstream dependencies:**
- Stripe API status: https://status.stripe.com
- Database CPU: [dashboard panel](https://grafana.example.com/d/db?from=now-1h)
- Redis hit rate: [panel](https://grafana.example.com/d/redis?panel=4)
## Mitigations (try in order)
### 1. If recent deploy correlates with onset
```bash
kubectl rollout undo deployment/payment-service -n payments
``` 60s অপেক্ষা করুন। dashboard-এ latency কমেছে কিনা নিশ্চিত করুন।
2. If database is the bottleneck (CPU > 80%)
# Failover to read replica for non-critical reads
kubectl patch configmap payment-config \
--type merge \
-p '{"data":{"DB_READ_FROM_REPLICA":"true"}}'
kubectl rollout restart deployment/payment-service 3. If Stripe is the bottleneck
degraded mode চালু করুন (cash-on-delivery only):
kubectl patch configmap payment-config \
--type merge \
-p '{"data":{"DEGRADED_MODE":"true"}}' Escalation
- After 15 min unresolved → page payments-secondary
- After 30 min unresolved → page engineering manager
- If revenue > $5k/min impact → page VP Eng
Last reviewed
2026-04-15 by @alice
এর বৈশিষ্ট্যগুলো: deep link (শুধু "গিয়ে Grafana দেখো" নয়), copy-paste করার মতো command, আর least-risky থেকে most-risky পর্যন্ত সাজানো mitigation।
## Pre-launch monitoring shift
SEV-critical launch-এর জন্য, প্রথম 48 ঘণ্টা extra eyes-এর ব্যবস্থা রাখুন। launch ops-এ এটাই একমাত্র সবচেয়ে বেশি leverage-দেওয়া tradition।
```typescript
// Launch monitoring shift schedule
const launchShift = {
service: "new-checkout-flow",
launchTime: "2026-05-15 10:00 PT",
tier1: {
duration: "T+0 to T+4h",
staff: ["author", "reviewer", "on-call SRE"],
activity: "Active monitoring; 5-min dashboard checks",
},
tier2: {
duration: "T+4h to T+24h",
staff: ["on-call SRE"],
activity: "Hourly dashboard checks; lower threshold pages",
},
tier3: {
duration: "T+24h to T+48h",
staff: ["on-call SRE"],
activity: "Normal on-call; launch tag still active for prioritization",
},
}; Annual recertification
PRR এককালীন জিনিস নয়। Service drift করে। বার্ষিক recertification:
# A real CI job that checks PRR compliance
sre-prr-check --service payment-service
# Sample output:
# ✓ SLO defined and recording rules active
# ✓ Burn-rate alerts wired
# ✗ Dashboard returned 404 (was deleted in Grafana cleanup)
# ✓ Runbook URL responsive
# ✗ Last DR drill: 2024-11 (>365 days ago)
# ✗ Backup restore last tested: never
#
# Service payment-service: 4/8 PRR items compliant
# Status: NEEDS RECERTIFICATION recertification ব্যর্থ হওয়া মানে SRE engineering manager-এর কাছে escalate করে। শাস্তিমূলক কিছু নয় — এটা সেই back-pressure যেটা production-কে নীরবে পচে যাওয়া থেকে ঠেকায়।
আপডেটেড থাকুন
- Google SRE Book — Launch Coordination Engineering — মূল PRR checklist
- AWS Well-Architected Framework — pillars + review process
- Google Cloud Architecture Framework — AWS WA-এর প্রতিরূপ
- 12-factor app — service hygiene-এর জন্য pre-PRR baseline
মূল কথাগুলো
- PRR সবচেয়ে বেশি outage-এর কারণ ধরে ফেলে incident-এর সময়ের চেয়ে 10x কম খরচে
- checklist মুখস্থ নয়, চেক করতে হবে — লিখে রাখুন
- baseline-টা Terraform-এ bake করুন — সঠিক পথই একমাত্র পথ হয়ে ওঠে
- Runbook-এ deep link আর copy-paste করার মতো command লাগে, অস্পষ্ট ইশারা নয়
- প্রতি বছর recertify করুন — service drift করে, drift outage ঘটায়, drift প্রতিরোধযোগ্য