Posts

Introduction: The Memory Wall Modern LLMs can process context windows of 100K+ tokens. But there’s a hidden cost: the KV cache. As context grows, the memory required to store key-value pairs in attention explodes quadratically. This creates a bottleneck: Memory: KV cache can consume 10-100× more memory than model weights Bandwidth: Moving KV cache data becomes the primary latency source Cost: Serving long-context models requires expensive high-memory GPUs Two innovations address this: Grouped Query Attention (GQA) and Multi-Head Latent Attention (MLA). They reduce KV cache size by 4-8× while maintaining quality. ...

← Login Counter | Series Overview | Sieve of Eratosthenes → The Problem: Computing Pi by Throwing Darts Imagine a square dartboard with a circle inscribed inside it. Throw random darts at the square. The ratio of darts landing inside the circle to total darts thrown approaches π/4. Why? Mathematics: Square side length: 2 (from -1 to 1) Square area: 4 Circle radius: 1 Circle area: π × 1² = π Ratio: π/4 Throw 1 million darts, multiply by 4, and you’ve estimated π. More darts = better estimate. This is Monte Carlo simulation: using randomness to solve deterministic problems. ...

Backend Communication Current: WebSockets Server-Sent Events All Posts gRPC Streaming What are WebSockets? WebSockets provide full-duplex, bidirectional communication channels over a single TCP connection. Unlike HTTP’s request-response model, WebSockets enable both client and server to send messages independently at any time, making them ideal for real-time applications. ...

Introduction: The Quadratic Bottleneck Transformers revolutionized AI, but they have a fundamental flaw: quadratic scaling. Processing a sequence of length n requires O(n²) operations due to self-attention. Every token attends to every other token, creating an all-to-all comparison: Context length: 1K 10K 100K 1M Operations: 1M 100M 10B 1T Time (relative): 1× 100× 10,000× 1,000,000× This makes long-context processing prohibitively expensive. Enter State Space Models (SSMs), specifically Mamba: a new architecture that processes sequences in linear time O(n) while maintaining long-range dependencies. ...

← Sieve of Eratosthenes | Series Overview | Collatz Explorer → The Problem: Rendering Fractals in Parallel The Mandelbrot set is defined by a simple iterative formula: Start with z = 0 Repeatedly compute z = z² + c If |z| exceeds 2, the point escapes (not in the set) Color each pixel by iteration count The beauty: Each pixel is completely independent. Perfect for parallelism! The challenge: Some pixels escape in 5 iterations, others take 1000+. This creates load imbalance-some workers finish instantly while others grind away. ...

Backend Communication Current: Server-Sent Events HTTP Polling Patterns All Posts WebSockets What are Server-Sent Events (SSE)? Server-Sent Events (SSE) is a server push technology that enables servers to push real-time updates to clients over a single HTTP connection. Unlike WebSockets, SSE is uni-directional (server → client only) and uses the standard HTTP protocol, making it simpler and more reliable through proxies and firewalls. ...

Go Architecture Patterns Series: ← Previous: Modular Monolith | Series Overview | Next: Event-Driven Architecture → What is Microservices Architecture? Microservices Architecture is an approach where an application is composed of small, independent services that communicate over a network. Each service is self-contained, owns its data, and can be deployed, scaled, and updated independently. Key Principles: Service Independence: Each service is deployed and scaled independently Single Responsibility: Each service handles a specific business capability Decentralized Data: Each service owns its database API-First Design: Services communicate through well-defined APIs Resilience: Services handle failures gracefully Technology Diversity: Services can use different technologies Architecture Overview %%{init: {'theme':'dark', 'themeVariables': {'primaryTextColor':'#e5e7eb','secondaryTextColor':'#e5e7eb','tertiaryTextColor':'#e5e7eb','textColor':'#e5e7eb','nodeTextColor':'#e5e7eb','edgeLabelText':'#e5e7eb','clusterTextColor':'#e5e7eb','actorTextColor':'#e5e7eb'}}}%% graph TB subgraph "Client Layer" Client[Web/Mobile Client] end subgraph "API Gateway" Gateway[API Gateway / Load Balancer] end subgraph "Service Mesh" UserService[User Service] ProductService[Product Service] OrderService[Order Service] PaymentService[Payment Service] NotificationService[Notification Service] end subgraph "Data Layer" UserDB[(User DB)] ProductDB[(Product DB)] OrderDB[(Order DB)] PaymentDB[(Payment DB)] end subgraph "Infrastructure" MessageBroker[Message Broker] ServiceRegistry[Service Discovery] ConfigServer[Config Server] end Client --> Gateway Gateway --> UserService Gateway --> ProductService Gateway --> OrderService Gateway --> PaymentService UserService --> UserDB ProductService --> ProductDB OrderService --> OrderDB PaymentService --> PaymentDB OrderService -.->|HTTP/gRPC| UserService OrderService -.->|HTTP/gRPC| ProductService OrderService -.->|HTTP/gRPC| PaymentService OrderService -.->|Async| MessageBroker PaymentService -.->|Async| MessageBroker NotificationService -.->|Subscribe| MessageBroker UserService -.-> ServiceRegistry ProductService -.-> ServiceRegistry OrderService -.-> ServiceRegistry style UserService fill:#1e3a5f,color:#fff style ProductService fill:#78350f,color:#fff style OrderService fill:#134e4a,color:#fff style PaymentService fill:#4c1d95,color:#fff style NotificationService fill:#4a1e3a,color:#fff Service Communication Patterns %%{init: {'theme':'dark', 'themeVariables': {'primaryTextColor':'#e5e7eb','secondaryTextColor':'#e5e7eb','tertiaryTextColor':'#e5e7eb','textColor':'#e5e7eb','nodeTextColor':'#e5e7eb','edgeLabelText':'#e5e7eb','clusterTextColor':'#e5e7eb','actorTextColor':'#e5e7eb'}}}%% sequenceDiagram participant Client participant Gateway participant OrderSvc as Order Service participant UserSvc as User Service participant ProductSvc as Product Service participant PaymentSvc as Payment Service participant Queue as Message Queue participant NotifySvc as Notification Service Client->>Gateway: Create Order Request Gateway->>OrderSvc: POST /orders OrderSvc->>UserSvc: GET /users/{id} UserSvc-->>OrderSvc: User Data OrderSvc->>ProductSvc: GET /products/{id} ProductSvc-->>OrderSvc: Product Data OrderSvc->>ProductSvc: POST /products/reserve ProductSvc-->>OrderSvc: Stock Reserved OrderSvc->>PaymentSvc: POST /payments PaymentSvc-->>OrderSvc: Payment Success OrderSvc->>Queue: Publish OrderCreated Event Queue->>NotifySvc: OrderCreated Event NotifySvc->>NotifySvc: Send Email/SMS OrderSvc-->>Gateway: Order Created Gateway-->>Client: Response Real-World Use Cases E-commerce Platforms: Amazon, eBay with separate services for products, orders, payments Streaming Services: Netflix with services for recommendations, playback, billing Ride-Sharing Apps: Uber with services for riders, drivers, payments, routing Financial Systems: Banking apps with separate services for accounts, transactions, loans Social Media: Twitter with services for posts, timelines, notifications, messages Cloud Platforms: AWS-like platforms with independent service offerings Microservices Implementation Project Structure (Multi-Repository) microservices/ ├── user-service/ │ ├── cmd/ │ │ └── server/ │ │ └── main.go │ ├── internal/ │ │ ├── domain/ │ │ ├── handlers/ │ │ ├── repository/ │ │ └── service/ │ ├── proto/ │ │ └── user.proto │ └── go.mod ├── product-service/ │ ├── cmd/ │ │ └── server/ │ │ └── main.go │ ├── internal/ │ │ ├── domain/ │ │ ├── handlers/ │ │ ├── repository/ │ │ └── service/ │ └── go.mod ├── order-service/ │ ├── cmd/ │ │ └── server/ │ │ └── main.go │ ├── internal/ │ │ ├── domain/ │ │ ├── handlers/ │ │ ├── repository/ │ │ ├── service/ │ │ └── clients/ │ └── go.mod └── api-gateway/ ├── cmd/ │ └── server/ │ └── main.go └── go.mod Service 1: User Service // user-service/internal/domain/user.go package domain import ( "context" "errors" "time" ) type User struct { ID string `json:"id"` Email string `json:"email"` Name string `json:"name"` Active bool `json:"active"` CreatedAt time.Time `json:"created_at"` } var ( ErrUserNotFound = errors.New("user not found") ErrUserExists = errors.New("user already exists") ) type Repository interface { Create(ctx context.Context, user *User) error GetByID(ctx context.Context, id string) (*User, error) GetByEmail(ctx context.Context, email string) (*User, error) Update(ctx context.Context, user *User) error } // user-service/internal/service/user_service.go package service import ( "context" "fmt" "time" "user-service/internal/domain" ) type UserService struct { repo domain.Repository } func NewUserService(repo domain.Repository) *UserService { return &UserService{repo: repo} } func (s *UserService) CreateUser(ctx context.Context, email, name string) (*domain.User, error) { existing, _ := s.repo.GetByEmail(ctx, email) if existing != nil { return nil, domain.ErrUserExists } user := &domain.User{ ID: generateID(), Email: email, Name: name, Active: true, CreatedAt: time.Now(), } if err := s.repo.Create(ctx, user); err != nil { return nil, fmt.Errorf("failed to create user: %w", err) } return user, nil } func (s *UserService) GetUser(ctx context.Context, id string) (*domain.User, error) { return s.repo.GetByID(ctx, id) } func (s *UserService) ValidateUser(ctx context.Context, id string) (bool, error) { user, err := s.repo.GetByID(ctx, id) if err != nil { return false, err } return user.Active, nil } func generateID() string { return fmt.Sprintf("user_%d", time.Now().UnixNano()) } // user-service/internal/handlers/http_handler.go package handlers import ( "encoding/json" "net/http" "github.com/gorilla/mux" "user-service/internal/service" ) type HTTPHandler struct { service *service.UserService } func NewHTTPHandler(service *service.UserService) *HTTPHandler { return &HTTPHandler{service: service} } type CreateUserRequest struct { Email string `json:"email"` Name string `json:"name"` } func (h *HTTPHandler) CreateUser(w http.ResponseWriter, r *http.Request) { var req CreateUserRequest if err := json.NewDecoder(r.Body).Decode(&req); err != nil { respondError(w, http.StatusBadRequest, "invalid request") return } user, err := h.service.CreateUser(r.Context(), req.Email, req.Name) if err != nil { respondError(w, http.StatusBadRequest, err.Error()) return } respondJSON(w, http.StatusCreated, user) } func (h *HTTPHandler) GetUser(w http.ResponseWriter, r *http.Request) { vars := mux.Vars(r) id := vars["id"] user, err := h.service.GetUser(r.Context(), id) if err != nil { respondError(w, http.StatusNotFound, "user not found") return } respondJSON(w, http.StatusOK, user) } func (h *HTTPHandler) ValidateUser(w http.ResponseWriter, r *http.Request) { vars := mux.Vars(r) id := vars["id"] valid, err := h.service.ValidateUser(r.Context(), id) if err != nil { respondError(w, http.StatusNotFound, "user not found") return } respondJSON(w, http.StatusOK, map[string]bool{"valid": valid}) } func respondJSON(w http.ResponseWriter, status int, data interface{}) { w.Header().Set("Content-Type", "application/json") w.WriteHeader(status) json.NewEncoder(w).Encode(data) } func respondError(w http.ResponseWriter, status int, message string) { respondJSON(w, status, map[string]string{"error": message}) } // user-service/cmd/server/main.go package main import ( "database/sql" "log" "net/http" "os" "github.com/gorilla/mux" _ "github.com/lib/pq" "user-service/internal/handlers" "user-service/internal/repository" "user-service/internal/service" ) func main() { dbURL := os.Getenv("DATABASE_URL") if dbURL == "" { dbURL = "postgres://user:pass@localhost/users?sslmode=disable" } db, err := sql.Open("postgres", dbURL) if err != nil { log.Fatal(err) } defer db.Close() repo := repository.NewPostgresRepository(db) svc := service.NewUserService(repo) handler := handlers.NewHTTPHandler(svc) router := mux.NewRouter() router.HandleFunc("/users", handler.CreateUser).Methods("POST") router.HandleFunc("/users/{id}", handler.GetUser).Methods("GET") router.HandleFunc("/users/{id}/validate", handler.ValidateUser).Methods("GET") router.HandleFunc("/health", func(w http.ResponseWriter, r *http.Request) { w.WriteHeader(http.StatusOK) w.Write([]byte("OK")) }) port := os.Getenv("PORT") if port == "" { port = "8081" } log.Printf("User service starting on port %s", port) if err := http.ListenAndServe(":"+port, router); err != nil { log.Fatal(err) } } Service 2: Product Service // product-service/internal/domain/product.go package domain import ( "context" "errors" "time" ) type Product struct { ID string `json:"id"` Name string `json:"name"` Description string `json:"description"` Price float64 `json:"price"` Stock int `json:"stock"` CreatedAt time.Time `json:"created_at"` } var ( ErrProductNotFound = errors.New("product not found") ErrInsufficientStock = errors.New("insufficient stock") ) type Repository interface { Create(ctx context.Context, product *Product) error GetByID(ctx context.Context, id string) (*Product, error) Update(ctx context.Context, product *Product) error ReserveStock(ctx context.Context, id string, quantity int) error } // product-service/internal/service/product_service.go package service import ( "context" "fmt" "time" "product-service/internal/domain" ) type ProductService struct { repo domain.Repository } func NewProductService(repo domain.Repository) *ProductService { return &ProductService{repo: repo} } func (s *ProductService) CreateProduct(ctx context.Context, name, desc string, price float64, stock int) (*domain.Product, error) { product := &domain.Product{ ID: generateID(), Name: name, Description: desc, Price: price, Stock: stock, CreatedAt: time.Now(), } if err := s.repo.Create(ctx, product); err != nil { return nil, fmt.Errorf("failed to create product: %w", err) } return product, nil } func (s *ProductService) GetProduct(ctx context.Context, id string) (*domain.Product, error) { return s.repo.GetByID(ctx, id) } func (s *ProductService) ReserveStock(ctx context.Context, id string, quantity int) error { product, err := s.repo.GetByID(ctx, id) if err != nil { return err } if product.Stock < quantity { return domain.ErrInsufficientStock } return s.repo.ReserveStock(ctx, id, quantity) } func generateID() string { return fmt.Sprintf("product_%d", time.Now().UnixNano()) } // product-service/cmd/server/main.go package main import ( "database/sql" "encoding/json" "log" "net/http" "os" "github.com/gorilla/mux" _ "github.com/lib/pq" "product-service/internal/repository" "product-service/internal/service" ) func main() { dbURL := os.Getenv("DATABASE_URL") if dbURL == "" { dbURL = "postgres://user:pass@localhost/products?sslmode=disable" } db, err := sql.Open("postgres", dbURL) if err != nil { log.Fatal(err) } defer db.Close() repo := repository.NewPostgresRepository(db) svc := service.NewProductService(repo) router := mux.NewRouter() router.HandleFunc("/products/{id}", func(w http.ResponseWriter, r *http.Request) { vars := mux.Vars(r) product, err := svc.GetProduct(r.Context(), vars["id"]) if err != nil { http.Error(w, err.Error(), http.StatusNotFound) return } json.NewEncoder(w).Encode(product) }).Methods("GET") router.HandleFunc("/products/reserve", func(w http.ResponseWriter, r *http.Request) { var req struct { ProductID string `json:"product_id"` Quantity int `json:"quantity"` } if err := json.NewDecoder(r.Body).Decode(&req); err != nil { http.Error(w, err.Error(), http.StatusBadRequest) return } if err := svc.ReserveStock(r.Context(), req.ProductID, req.Quantity); err != nil { http.Error(w, err.Error(), http.StatusBadRequest) return } w.WriteHeader(http.StatusOK) json.NewEncoder(w).Encode(map[string]string{"status": "reserved"}) }).Methods("POST") router.HandleFunc("/health", func(w http.ResponseWriter, r *http.Request) { w.WriteHeader(http.StatusOK) }) port := os.Getenv("PORT") if port == "" { port = "8082" } log.Printf("Product service starting on port %s", port) if err := http.ListenAndServe(":"+port, router); err != nil { log.Fatal(err) } } Service 3: Order Service (Orchestrator) // order-service/internal/clients/user_client.go package clients import ( "context" "encoding/json" "fmt" "net/http" "time" ) type UserClient struct { baseURL string httpClient *http.Client } func NewUserClient(baseURL string) *UserClient { return &UserClient{ baseURL: baseURL, httpClient: &http.Client{ Timeout: 5 * time.Second, }, } } func (c *UserClient) ValidateUser(ctx context.Context, userID string) (bool, error) { url := fmt.Sprintf("%s/users/%s/validate", c.baseURL, userID) req, err := http.NewRequestWithContext(ctx, "GET", url, nil) if err != nil { return false, err } resp, err := c.httpClient.Do(req) if err != nil { return false, fmt.Errorf("failed to call user service: %w", err) } defer resp.Body.Close() if resp.StatusCode != http.StatusOK { return false, fmt.Errorf("user service returned status %d", resp.StatusCode) } var result struct { Valid bool `json:"valid"` } if err := json.NewDecoder(resp.Body).Decode(&result); err != nil { return false, err } return result.Valid, nil } // order-service/internal/clients/product_client.go package clients import ( "bytes" "context" "encoding/json" "fmt" "net/http" "time" ) type Product struct { ID string `json:"id"` Name string `json:"name"` Price float64 `json:"price"` Stock int `json:"stock"` } type ProductClient struct { baseURL string httpClient *http.Client } func NewProductClient(baseURL string) *ProductClient { return &ProductClient{ baseURL: baseURL, httpClient: &http.Client{ Timeout: 5 * time.Second, }, } } func (c *ProductClient) GetProduct(ctx context.Context, productID string) (*Product, error) { url := fmt.Sprintf("%s/products/%s", c.baseURL, productID) req, err := http.NewRequestWithContext(ctx, "GET", url, nil) if err != nil { return nil, err } resp, err := c.httpClient.Do(req) if err != nil { return nil, fmt.Errorf("failed to call product service: %w", err) } defer resp.Body.Close() if resp.StatusCode != http.StatusOK { return nil, fmt.Errorf("product not found") } var product Product if err := json.NewDecoder(resp.Body).Decode(&product); err != nil { return nil, err } return &product, nil } func (c *ProductClient) ReserveStock(ctx context.Context, productID string, quantity int) error { url := fmt.Sprintf("%s/products/reserve", c.baseURL) reqBody := map[string]interface{}{ "product_id": productID, "quantity": quantity, } body, err := json.Marshal(reqBody) if err != nil { return err } req, err := http.NewRequestWithContext(ctx, "POST", url, bytes.NewBuffer(body)) if err != nil { return err } req.Header.Set("Content-Type", "application/json") resp, err := c.httpClient.Do(req) if err != nil { return fmt.Errorf("failed to reserve stock: %w", err) } defer resp.Body.Close() if resp.StatusCode != http.StatusOK { return fmt.Errorf("failed to reserve stock: status %d", resp.StatusCode) } return nil } // order-service/internal/service/order_service.go package service import ( "context" "fmt" "time" "order-service/internal/clients" "order-service/internal/domain" ) type OrderService struct { repo domain.Repository userClient *clients.UserClient productClient *clients.ProductClient } func NewOrderService( repo domain.Repository, userClient *clients.UserClient, productClient *clients.ProductClient, ) *OrderService { return &OrderService{ repo: repo, userClient: userClient, productClient: productClient, } } func (s *OrderService) CreateOrder(ctx context.Context, userID string, items []domain.OrderItem) (*domain.Order, error) { // Validate user via User Service valid, err := s.userClient.ValidateUser(ctx, userID) if err != nil { return nil, fmt.Errorf("failed to validate user: %w", err) } if !valid { return nil, fmt.Errorf("user is not valid") } // Validate products and calculate total var total float64 for i, item := range items { product, err := s.productClient.GetProduct(ctx, item.ProductID) if err != nil { return nil, fmt.Errorf("failed to get product: %w", err) } items[i].Price = product.Price total += product.Price * float64(item.Quantity) } // Reserve stock via Product Service for _, item := range items { if err := s.productClient.ReserveStock(ctx, item.ProductID, item.Quantity); err != nil { return nil, fmt.Errorf("failed to reserve stock: %w", err) } } order := &domain.Order{ ID: generateID(), UserID: userID, Items: items, Total: total, Status: "pending", CreatedAt: time.Now(), } if err := s.repo.Create(ctx, order); err != nil { return nil, fmt.Errorf("failed to create order: %w", err) } return order, nil } func generateID() string { return fmt.Sprintf("order_%d", time.Now().UnixNano()) } // order-service/cmd/server/main.go package main import ( "database/sql" "encoding/json" "log" "net/http" "os" "github.com/gorilla/mux" _ "github.com/lib/pq" "order-service/internal/clients" "order-service/internal/domain" "order-service/internal/repository" "order-service/internal/service" ) func main() { dbURL := os.Getenv("DATABASE_URL") if dbURL == "" { dbURL = "postgres://user:pass@localhost/orders?sslmode=disable" } userServiceURL := os.Getenv("USER_SERVICE_URL") if userServiceURL == "" { userServiceURL = "http://localhost:8081" } productServiceURL := os.Getenv("PRODUCT_SERVICE_URL") if productServiceURL == "" { productServiceURL = "http://localhost:8082" } db, err := sql.Open("postgres", dbURL) if err != nil { log.Fatal(err) } defer db.Close() repo := repository.NewPostgresRepository(db) userClient := clients.NewUserClient(userServiceURL) productClient := clients.NewProductClient(productServiceURL) svc := service.NewOrderService(repo, userClient, productClient) router := mux.NewRouter() router.HandleFunc("/orders", func(w http.ResponseWriter, r *http.Request) { var req struct { UserID string `json:"user_id"` Items []domain.OrderItem `json:"items"` } if err := json.NewDecoder(r.Body).Decode(&req); err != nil { http.Error(w, err.Error(), http.StatusBadRequest) return } order, err := svc.CreateOrder(r.Context(), req.UserID, req.Items) if err != nil { http.Error(w, err.Error(), http.StatusBadRequest) return } w.Header().Set("Content-Type", "application/json") w.WriteHeader(http.StatusCreated) json.NewEncoder(w).Encode(order) }).Methods("POST") router.HandleFunc("/health", func(w http.ResponseWriter, r *http.Request) { w.WriteHeader(http.StatusOK) }) port := os.Getenv("PORT") if port == "" { port = "8083" } log.Printf("Order service starting on port %s", port) if err := http.ListenAndServe(":"+port, router); err != nil { log.Fatal(err) } } Docker Compose Setup version: '3.8' services: user-service: build: ./user-service ports: - "8081:8081" environment: - DATABASE_URL=postgres://user:pass@user-db:5432/users?sslmode=disable - PORT=8081 depends_on: - user-db product-service: build: ./product-service ports: - "8082:8082" environment: - DATABASE_URL=postgres://user:pass@product-db:5432/products?sslmode=disable - PORT=8082 depends_on: - product-db order-service: build: ./order-service ports: - "8083:8083" environment: - DATABASE_URL=postgres://user:pass@order-db:5432/orders?sslmode=disable - USER_SERVICE_URL=http://user-service:8081 - PRODUCT_SERVICE_URL=http://product-service:8082 - PORT=8083 depends_on: - order-db - user-service - product-service user-db: image: postgres:15 environment: - POSTGRES_USER=user - POSTGRES_PASSWORD=pass - POSTGRES_DB=users product-db: image: postgres:15 environment: - POSTGRES_USER=user - POSTGRES_PASSWORD=pass - POSTGRES_DB=products order-db: image: postgres:15 environment: - POSTGRES_USER=user - POSTGRES_PASSWORD=pass - POSTGRES_DB=orders Best Practices Service Boundaries: Define clear service boundaries based on business capabilities API Contracts: Use API versioning and maintain backward compatibility Service Discovery: Implement service registry for dynamic service location Circuit Breakers: Prevent cascading failures with circuit breaker pattern Distributed Tracing: Implement tracing to debug cross-service calls Health Checks: Provide health endpoints for monitoring Configuration Management: Externalize configuration Security: Implement service-to-service authentication Common Pitfalls Distributed Monolith: Services too tightly coupled, defeating the purpose Chatty Services: Too many synchronous calls between services Shared Database: Multiple services accessing the same database Ignoring Network Failures: Not handling network errors gracefully No Service Versioning: Breaking changes without versioning Data Consistency Issues: Not handling eventual consistency Over-Engineering: Creating too many small services When to Use Microservices Architecture Use When: ...

← Ticket Seller | Series Overview | Monte Carlo Pi → The Problem: Counting What You Can’t See Count concurrent users. On login: increment. On logout: decrement. Simple, right? Now add reality: The counter is distributed across multiple servers A user’s session times out on one server but they’re active on another The increment message arrives after the decrement message Network partitions split your cluster Servers crash mid-operation Suddenly, this “simple” counter becomes a distributed systems nightmare. Welcome to distributed counting, where even addition is hard. ...

Introduction Release management is the process of planning, scheduling, and controlling software releases through different stages and environments. It ensures that software is released reliably, predictably, and with minimal disruption. This guide visualizes key release management concepts: Semantic Versioning: Deciding when to bump major, minor, or patch versions Release Train: Structured release cadence with quality gates Hotfix Process: Fast-track critical fixes to production Release Checklist: Ensuring nothing is missed during deployment Environment Promotion: Moving code through dev, staging, and production Part 1: Semantic Versioning Decision Tree Understanding Version Numbers: MAJOR.MINOR.PATCH Semantic versioning (SemVer) uses a three-part version number: MAJOR.MINOR.PATCH ...

Introduction CI/CD (Continuous Integration/Continuous Deployment) automates the software delivery process from code commit to production deployment. This automation reduces manual errors, speeds up releases, and improves software quality. This guide visualizes the complete CI/CD pipeline: Code Commit: Developer pushes code Continuous Integration: Automated testing and building Continuous Deployment: Automated deployment to production Quality Gates: Checkpoints ensuring code quality Rollback Mechanisms: Handling deployment failures Part 1: Complete CI/CD Pipeline Overview End-to-End Flow %%{init: {'theme':'dark', 'themeVariables': {'primaryTextColor':'#e5e7eb','secondaryTextColor':'#e5e7eb','tertiaryTextColor':'#e5e7eb','textColor':'#e5e7eb','nodeTextColor':'#e5e7eb','edgeLabelText':'#e5e7eb','clusterTextColor':'#e5e7eb','actorTextColor':'#e5e7eb'}}}%% flowchart TD Start([Developer writes codecommits changes]) --> Push[git push origin main] Push --> Webhook[Git Provider WebhookTriggers CI/CD pipeline] Webhook --> Checkout[Stage 1: CheckoutClone repositoryFetch dependencies] Checkout --> Lint[Stage 2: LintCheck code styleESLint, Prettier, golangci-lint] Lint --> LintResult{Lintingpassed?} LintResult -->|No| LintFail[❌ Pipeline FailedNotify developerFix linting errors] LintResult -->|Yes| UnitTest[Stage 3: Unit TestsRun test suiteGenerate coverage report] UnitTest --> TestResult{Testspassed?} TestResult -->|No| TestFail[❌ Pipeline FailedSome tests failedCoverage too low] TestResult -->|Yes| Build[Stage 4: BuildCompile applicationBuild Docker image] Build --> BuildResult{Buildsuccessful?} BuildResult -->|No| BuildFail[❌ Pipeline FailedBuild errorsDependency issues] BuildResult -->|Yes| IntegTest[Stage 5: Integration TestsTest with real dependenciesDatabase, APIs, etc.] IntegTest --> IntegResult{Integrationtests passed?} IntegResult -->|No| IntegFail[❌ Pipeline FailedIntegration issuesService communication errors] IntegResult -->|Yes| Security[Stage 6: Security ScanScan for vulnerabilitiesOWASP, Snyk, Trivy] Security --> SecResult{Securitychecks passed?} SecResult -->|No| SecFail[❌ Pipeline FailedSecurity vulnerabilities foundFix before deploying] SecResult -->|Yes| Push2Registry[Stage 7: Push ImageTag: myapp:abc123Push to container registry] Push2Registry --> DeployStaging[Stage 8: Deploy to Stagingkubectl apply -f staging/Run smoke tests] DeployStaging --> SmokeTest[Stage 9: Smoke TestsTest critical pathsHealth checksBasic functionality] SmokeTest --> SmokeResult{Smoke testspassed?} SmokeResult -->|No| StagingFail[❌ Pipeline FailedStaging deployment issuesRollback staging] SmokeResult -->|Yes| Approval{Manualapprovalrequired?} Approval -->|Yes| WaitApproval[⏸️ Waiting for ApprovalNotify team leadReview changes] WaitApproval --> ApprovalDecision{Approved?} ApprovalDecision -->|No| Rejected[❌ Deployment RejectedNot ready for production] ApprovalDecision -->|Yes| DeployProd Approval -->|No| DeployProd[Stage 10: Deploy to ProductionRolling updateOr blue-green deployment] DeployProd --> ProdHealth{Productionhealthy?} ProdHealth -->|No| AutoRollback[❌ Auto-RollbackRevert to previous versionAlert on-call team] ProdHealth -->|Yes| Success[✅ Deployment Successful!Monitor metricsNotify teamUpdate status] style LintFail fill:#7f1d1d,stroke:#ef4444 style TestFail fill:#7f1d1d,stroke:#ef4444 style BuildFail fill:#7f1d1d,stroke:#ef4444 style IntegFail fill:#7f1d1d,stroke:#ef4444 style SecFail fill:#7f1d1d,stroke:#ef4444 style StagingFail fill:#7f1d1d,stroke:#ef4444 style AutoRollback fill:#7f1d1d,stroke:#ef4444 style Success fill:#064e3b,stroke:#10b981 style WaitApproval fill:#78350f,stroke:#f59e0b Part 2: Continuous Integration (CI) Stages CI Pipeline Detailed Flow %%{init: {'theme':'dark', 'themeVariables': {'primaryTextColor':'#e5e7eb','secondaryTextColor':'#e5e7eb','tertiaryTextColor':'#e5e7eb','textColor':'#e5e7eb','nodeTextColor':'#e5e7eb','edgeLabelText':'#e5e7eb','clusterTextColor':'#e5e7eb','actorTextColor':'#e5e7eb'}}}%% sequenceDiagram participant Dev as Developer participant Git as Git Repository participant CI as CI Server participant Docker as Docker Registry participant Notify as Slack/Email Dev->>Git: git push origin feature/new-api Note over Git: Webhook triggeredon push event Git->>CI: Trigger pipeline:Branch: feature/new-apiCommit: abc123Author: [email protected] CI->>CI: Create build environmentUbuntu 22.04 container CI->>Git: git clone --depth 1Checkout abc123 Note over CI: Stage 1: Setup CI->>CI: Install dependenciesnpm installgo mod download Note over CI: Stage 2: Code Quality CI->>CI: Run lintereslint src/golangci-lint run alt Linting Failed CI->>Notify: ❌ Linting failed26 issues foundFix formatting CI-->>Dev: Pipeline failed end Note over CI: Stage 3: Unit Testing CI->>CI: Run unit testsnpm testgo test ./... CI->>CI: Generate coverage reportCoverage: 87% alt Tests Failed or Low Coverage CI->>Notify: ❌ Tests failed5 tests failingCoverage: 72% < 80% CI-->>Dev: Pipeline failed end Note over CI: Stage 4: Build CI->>CI: Build applicationnpm run buildgo build -o app CI->>CI: Build Docker imagedocker build -t myapp:abc123 alt Build Failed CI->>Notify: ❌ Build failedCompilation errors CI-->>Dev: Pipeline failed end Note over CI: Stage 5: Integration Tests CI->>CI: Start test dependenciesdocker-compose up -dpostgres, redis CI->>CI: Run integration testsTest database connectionsTest API endpoints CI->>CI: Stop test servicesdocker-compose down alt Integration Tests Failed CI->>Notify: ❌ Integration tests failedDatabase connection timeout CI-->>Dev: Pipeline failed end Note over CI: Stage 6: Security Scanning CI->>CI: Scan dependenciesnpm auditsnyk test CI->>CI: Scan Docker imagetrivy image myapp:abc123 alt Security Issues Found CI->>Notify: ⚠️ Security issues3 high severity CVEsUpdate dependencies CI-->>Dev: Pipeline failed end Note over CI: All checks passed! ✓ CI->>Docker: docker push myapp:abc123Tag: myapp:latest Docker-->>CI: Image pushed successfully CI->>Notify: ✅ Build successful!Image: myapp:abc123Ready for deployment CI-->>Dev: Pipeline succeededDuration: 8m 32s GitHub Actions CI Configuration # .github/workflows/ci.yml name: CI Pipeline on: push: branches: [ main, develop ] pull_request: branches: [ main ] env: REGISTRY: ghcr.io IMAGE_NAME: ${{ github.repository }} jobs: # Job 1: Code Quality Checks lint: runs-on: ubuntu-latest steps: - uses: actions/checkout@v3 - name: Setup Node.js uses: actions/setup-node@v3 with: node-version: '18' cache: 'npm' - name: Install dependencies run: npm ci - name: Run ESLint run: npm run lint - name: Run Prettier run: npm run format:check # Job 2: Unit Tests test: runs-on: ubuntu-latest steps: - uses: actions/checkout@v3 - name: Setup Node.js uses: actions/setup-node@v3 with: node-version: '18' cache: 'npm' - name: Install dependencies run: npm ci - name: Run tests run: npm test -- --coverage - name: Check coverage threshold run: | COVERAGE=$(cat coverage/coverage-summary.json | jq '.total.lines.pct') if (( $(echo "$COVERAGE < 80" | bc -l) )); then echo "Coverage $COVERAGE% is below 80%" exit 1 fi - name: Upload coverage to Codecov uses: codecov/codecov-action@v3 # Job 3: Build build: runs-on: ubuntu-latest needs: [lint, test] # Wait for lint and test to pass steps: - uses: actions/checkout@v3 - name: Set up Docker Buildx uses: docker/setup-buildx-action@v2 - name: Log in to GitHub Container Registry uses: docker/login-action@v2 with: registry: ${{ env.REGISTRY }} username: ${{ github.actor }} password: ${{ secrets.GITHUB_TOKEN }} - name: Extract metadata id: meta uses: docker/metadata-action@v4 with: images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }} tags: | type=sha,prefix={{branch}}- type=ref,event=branch type=ref,event=pr - name: Build and push Docker image uses: docker/build-push-action@v4 with: context: . push: true tags: ${{ steps.meta.outputs.tags }} cache-from: type=gha cache-to: type=gha,mode=max # Job 4: Integration Tests integration-test: runs-on: ubuntu-latest needs: build services: postgres: image: postgres:15 env: POSTGRES_PASSWORD: postgres options: >- --health-cmd pg_isready --health-interval 10s --health-timeout 5s --health-retries 5 redis: image: redis:7 options: >- --health-cmd "redis-cli ping" --health-interval 10s --health-timeout 5s --health-retries 5 steps: - uses: actions/checkout@v3 - name: Setup Node.js uses: actions/setup-node@v3 with: node-version: '18' cache: 'npm' - name: Install dependencies run: npm ci - name: Run integration tests run: npm run test:integration env: DATABASE_URL: postgresql://postgres:postgres@localhost:5432/test REDIS_URL: redis://localhost:6379 # Job 5: Security Scan security: runs-on: ubuntu-latest needs: build steps: - uses: actions/checkout@v3 - name: Run npm audit run: npm audit --audit-level=high - name: Run Snyk security scan uses: snyk/actions/node@master env: SNYK_TOKEN: ${{ secrets.SNYK_TOKEN }} - name: Scan Docker image with Trivy uses: aquasecurity/trivy-action@master with: image-ref: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:${{ github.sha }} format: 'sarif' output: 'trivy-results.sarif' - name: Upload Trivy results to GitHub Security uses: github/codeql-action/upload-sarif@v2 with: sarif_file: 'trivy-results.sarif' Part 3: Continuous Deployment (CD) Stages Deployment Pipeline Flow %%{init: {'theme':'dark', 'themeVariables': {'primaryTextColor':'#e5e7eb','secondaryTextColor':'#e5e7eb','tertiaryTextColor':'#e5e7eb','textColor':'#e5e7eb','nodeTextColor':'#e5e7eb','edgeLabelText':'#e5e7eb','clusterTextColor':'#e5e7eb','actorTextColor':'#e5e7eb'}}}%% flowchart TD Start([CI Pipeline PassedImage ready: myapp:abc123]) --> DeployDecision{Whichbranch?} DeployDecision -->|feature/*| SkipDeploy[Skip deploymentCI only forfeature branches] DeployDecision -->|develop| DeployDev[Deploy to Dev EnvironmentNamespace: devAuto-deploy on push] DeployDecision -->|main| DeployStaging[Deploy to StagingNamespace: stagingAuto-deploy on push] DeployDev --> DevSmoke[Run smoke testsBasic health checks] DevSmoke --> DevDone[✅ Dev deployment complete] DeployStaging --> UpdateManifest[Update Kubernetes manifestsimage: myapp:abc123Apply configuration] UpdateManifest --> ApplyStaging[kubectl apply -f k8s/staging/Create/Update resourcesWait for rollout] ApplyStaging --> WaitReady{All podsready?} WaitReady -->|No timeout| CheckHealth[Check pod statuskubectl get pods -n staging] CheckHealth --> HealthStatus{Healthy?} HealthStatus -->|No| RollbackStaging[❌ Rollback stagingkubectl rollout undodeployment myapp -n staging] RollbackStaging --> NotifyFail[Notify team:Staging deployment failedCheck logs and fix] HealthStatus -->|Yes| StagingSmoke[Run staging smoke tests- Health endpoint- Critical API endpoints- Database connectivity] StagingSmoke --> SmokePass{Smoke testspassed?} SmokePass -->|No| RollbackStaging SmokePass -->|Yes| StagingReady[✅ Staging ReadyAll tests passedReady for production] StagingReady --> ApprovalGate{Require manualapproval?} ApprovalGate -->|Yes| WaitApproval[⏸️ Wait for approvalCreate deployment requestNotify reviewers] WaitApproval --> ReviewDecision{Approvedby reviewer?} ReviewDecision -->|No| Rejected[❌ Deployment rejectedFeedback providedMake changes] ReviewDecision -->|Yes| DeployProd ApprovalGate -->|No| DeployProd[Deploy to ProductionNamespace: productionStrategy: Rolling update] DeployProd --> BackupProd[Create backup:- Current deployment state- Database snapshot- Config backup] BackupProd --> ApplyProd[kubectl apply -f k8s/prod/Rolling update:maxSurge: 1maxUnavailable: 0] ApplyProd --> MonitorRollout[Monitor rollout statuskubectl rollout statusdeployment myapp -n production] MonitorRollout --> ProdHealth{All new podshealthy?} ProdHealth -->|No| AutoRollback[🚨 Auto-rollback triggeredkubectl rollout undoRestore previous version] AutoRollback --> AlertTeam[Alert on-call teamPagerDuty notificationProduction incident] ProdHealth -->|Yes| ProdMonitor[Monitor production metrics- Error rates- Latency- Business KPIs] ProdMonitor --> MetricsOK{Metricshealthy for10 minutes?} MetricsOK -->|No| AutoRollback MetricsOK -->|Yes| Complete[✅ Deployment Complete!Production healthyNew version liveUpdate status page] Complete --> CleanupOld[Cleanup old resourcesRemove old replica setsPrune old images] style SkipDeploy fill:#1e3a8a,stroke:#3b82f6 style WaitApproval fill:#78350f,stroke:#f59e0b style RollbackStaging fill:#7f1d1d,stroke:#ef4444 style AutoRollback fill:#7f1d1d,stroke:#ef4444 style Complete fill:#064e3b,stroke:#10b981 style DevDone fill:#064e3b,stroke:#10b981 Part 4: Quality Gates Quality Gate Decision Flow %%{init: {'theme':'dark', 'themeVariables': {'primaryTextColor':'#e5e7eb','secondaryTextColor':'#e5e7eb','tertiaryTextColor':'#e5e7eb','textColor':'#e5e7eb','nodeTextColor':'#e5e7eb','edgeLabelText':'#e5e7eb','clusterTextColor':'#e5e7eb','actorTextColor':'#e5e7eb'}}}%% flowchart TD Start([Code ready to deploy]) --> Gate1{Quality Gate 1:Code Quality} Gate1 --> CheckLint[Check LintingESLint, Prettier] Gate1 --> CheckComplexity[Check ComplexityCyclomatic complexity< 10 per function] Gate1 --> CheckDuplication[Check DuplicationCode duplication < 3%] CheckLint --> LintScore{Pass?} CheckComplexity --> ComplexScore{Pass?} CheckDuplication --> DupScore{Pass?} LintScore -->|No| Fail1[❌ Gate 1 Failed] ComplexScore -->|No| Fail1 DupScore -->|No| Fail1 LintScore -->|Yes| Gate2{Quality Gate 2:Testing} ComplexScore -->|Yes| Gate2 DupScore -->|Yes| Gate2 Gate2 --> CheckCoverage[Check CoverageLine coverage >= 80%Branch coverage >= 75%] Gate2 --> CheckTests[All Tests PassUnit + Integration] Gate2 --> CheckPerf[Performance TestsResponse time < baseline] CheckCoverage --> CovScore{Pass?} CheckTests --> TestScore{Pass?} CheckPerf --> PerfScore{Pass?} CovScore -->|No| Fail2[❌ Gate 2 Failed] TestScore -->|No| Fail2 PerfScore -->|No| Fail2 CovScore -->|Yes| Gate3{Quality Gate 3:Security} TestScore -->|Yes| Gate3 PerfScore -->|Yes| Gate3 Gate3 --> CheckVuln[Scan VulnerabilitiesNo high/critical CVEs] Gate3 --> CheckSecrets[Check for SecretsNo hardcoded credentials] Gate3 --> CheckDeps[Dependency CheckAll deps up-to-date] CheckVuln --> VulnScore{Pass?} CheckSecrets --> SecretScore{Pass?} CheckDeps --> DepScore{Pass?} VulnScore -->|No| Fail3[❌ Gate 3 Failed] SecretScore -->|No| Fail3 DepScore -->|No| Fail3 VulnScore -->|Yes| Gate4{Quality Gate 4:Production Readiness} SecretScore -->|Yes| Gate4 DepScore -->|Yes| Gate4 Gate4 --> CheckHealth[Health ChecksLiveness + Readiness] Gate4 --> CheckResources[Resource LimitsCPU + Memory defined] Gate4 --> CheckDocs[DocumentationREADME + API docs] CheckHealth --> HealthScore{Pass?} CheckResources --> ResScore{Pass?} CheckDocs --> DocScore{Pass?} HealthScore -->|No| Fail4[❌ Gate 4 Failed] ResScore -->|No| Fail4 DocScore -->|No| Fail4 HealthScore -->|Yes| AllGates[✅ All Quality Gates Passed!Ready for deployment] ResScore -->|Yes| AllGates DocScore -->|Yes| AllGates Fail1 --> Block[Block deploymentFix issues first] Fail2 --> Block Fail3 --> Block Fail4 --> Block style Fail1 fill:#7f1d1d,stroke:#ef4444 style Fail2 fill:#7f1d1d,stroke:#ef4444 style Fail3 fill:#7f1d1d,stroke:#ef4444 style Fail4 fill:#7f1d1d,stroke:#ef4444 style AllGates fill:#064e3b,stroke:#10b981 Part 5: GitLab CI/CD Example .gitlab-ci.yml Configuration # .gitlab-ci.yml stages: - lint - test - build - security - deploy-staging - deploy-production variables: DOCKER_DRIVER: overlay2 DOCKER_TLS_CERTDIR: "/certs" IMAGE_TAG: $CI_REGISTRY_IMAGE:$CI_COMMIT_SHORT_SHA # Template for Docker jobs .docker-login: &docker-login before_script: - docker login -u $CI_REGISTRY_USER -p $CI_REGISTRY_PASSWORD $CI_REGISTRY # Stage 1: Linting lint:code: stage: lint image: node:18 script: - npm ci - npm run lint - npm run format:check cache: paths: - node_modules/ # Stage 2: Testing test:unit: stage: test image: node:18 script: - npm ci - npm test -- --coverage - | COVERAGE=$(cat coverage/coverage-summary.json | jq '.total.lines.pct') if (( $(echo "$COVERAGE < 80" | bc -l) )); then echo "Coverage $COVERAGE% is below threshold" exit 1 fi coverage: '/Lines\s*:\s*(\d+\.\d+)%/' artifacts: reports: coverage_report: coverage_format: cobertura path: coverage/cobertura-coverage.xml test:integration: stage: test image: node:18 services: - name: postgres:15 alias: postgres - name: redis:7 alias: redis variables: DATABASE_URL: postgresql://postgres:postgres@postgres:5432/test REDIS_URL: redis://redis:6379 script: - npm ci - npm run test:integration # Stage 3: Build build:image: stage: build image: docker:24 services: - docker:24-dind <<: *docker-login script: - docker build -t $IMAGE_TAG . - docker push $IMAGE_TAG - docker tag $IMAGE_TAG $CI_REGISTRY_IMAGE:latest - docker push $CI_REGISTRY_IMAGE:latest only: - main - develop # Stage 4: Security Scanning security:scan: stage: security image: aquasec/trivy:latest script: - trivy image --severity HIGH,CRITICAL --exit-code 1 $IMAGE_TAG allow_failure: true security:sast: stage: security image: node:18 script: - npm audit --audit-level=high - npx snyk test --severity-threshold=high allow_failure: true # Stage 5: Deploy to Staging deploy:staging: stage: deploy-staging image: bitnami/kubectl:latest script: - kubectl config set-cluster k8s --server="$K8S_SERVER" - kubectl config set-credentials admin --token="$K8S_TOKEN" - kubectl config set-context default --cluster=k8s --user=admin - kubectl config use-context default - | kubectl set image deployment/myapp \ myapp=$IMAGE_TAG \ -n staging - kubectl rollout status deployment/myapp -n staging --timeout=5m - kubectl get pods -n staging environment: name: staging url: https://staging.example.com only: - main # Stage 6: Deploy to Production deploy:production: stage: deploy-production image: bitnami/kubectl:latest script: - kubectl config set-cluster k8s --server="$K8S_SERVER" - kubectl config set-credentials admin --token="$K8S_TOKEN" - kubectl config set-context default --cluster=k8s --user=admin - kubectl config use-context default - | kubectl set image deployment/myapp \ myapp=$IMAGE_TAG \ -n production - kubectl rollout status deployment/myapp -n production --timeout=10m - | # Check pod health READY=$(kubectl get deployment myapp -n production -o jsonpath='{.status.readyReplicas}') DESIRED=$(kubectl get deployment myapp -n production -o jsonpath='{.spec.replicas}') if [ "$READY" != "$DESIRED" ]; then echo "Deployment unhealthy: $READY/$DESIRED pods ready" kubectl rollout undo deployment/myapp -n production exit 1 fi environment: name: production url: https://example.com when: manual # Require manual approval only: - main Part 6: Pipeline Best Practices Pipeline Optimization Fast Feedback Loop: ...