- Set applied_rules capacity to 6 (maximum number of charging rules that can be applied)
- Set keys capacity to 100 in get_usage_history and get_all_accounts (matches SCAN batch size)
- Add comments explaining capacity choices for each Vec initialization
- Change AuthConfig::from_env() return type from Self to Result<Self, String>
- Remove default JWT_SECRET fallback, require environment variable to be set
- Add validation requiring JWT_SECRET to be at least 32 characters for security
- Replace expect() with ok_or_else in create_token timestamp calculation
- Replace expect() with map_err in DATABASE_URL environment variable loading
- Remove unused _account variable in calculate_usage_cost
- Replace HashMap::with_capacity with HashMap::new and use iterator chain in get_all_plans
- Add std::process::Command import and set current_dir to repo root in run_migrations
- Add comment explaining directory change to locate migrations folder
- Replace expect() with unwrap_or_else providing fallback JSON in SubscriberAccount and UsageEvent
- Add thiserror::Error derive to ChargingError enum
- Replace manual fmt::Display implementation with #[error] attributes on each variant
- Remove manual std::error::Error implementation (provided by thiserror)
- Remove std::fmt import as it's no longer needed
- Maintain same error message formats and functionality
- Remove clone() from rating plan parameter in add_rating_plan call
- Remove clone() calls from ChargingEngineClient and EbpfManager initialization
- Remove clone() from Redis client Arc wrapping
- Remove clone() calls from PacketGatewayConfig field assignments
- Replace expect() with proper error handling using map_err and anyhow in health check server
- Remove clone() from IP string in session data collection
- Add cargo-audit metadata section with placeholder for ignored advisories
- Forbid unsafe_code at workspace level
- Enable all and pedantic clippy warnings
- Deny unwrap_used, expect_used, and panic to enforce error handling
- Add thiserror dependency to charging-engine for improved error handling
- Add filepath import for path manipulation
- Calculate repository root path relative to carrier-connector location (../../)
- Set cmd.Dir to repository root before executing Goose command
- Add comment explaining directory change to locate migrations folder
- Ensure migrations can be found from carrier-connector's nested location
- Move all subscriber, payment method, transaction, and notification CRUD operations to crud.go
- Remove models import from database.go as it's no longer needed
- Keep database initialization, connection management, and IMSI allocation in database.go
- Maintain ListSubscribersRequest struct definition in crud.go alongside related functions
- Cast req.Status to string when passing to database.ListSubscribersRequest
- Ensure type compatibility between service and database layer status fields
- Move working-directory after env block in Run database migrations step
- Maintain same functionality while following standard YAML ordering convention
- Add subscribers table with IMSI, MSISDN, authentication keys, and eSIM profile fields
- Add service_plans table for data/voice/SMS limits and QoS parameters (ARPA, AMBR)
- Add sessions table for tracking PDU sessions with QoS and usage metrics
- Add usage_records table for billing and charging data
- Add users table with role-based access control fields
- Add roles, permissions, and role_permissions junction table for RBAC
- Add auth
- Change migrations path from './migrations' to 'migrations' in all test jobs
- Add explicit working-directory: . to migration steps for clarity
- Apply changes to api-server-test, charging-engine-test, and packet-gateway-test jobs
- Add explicit PATH configuration step to ensure goose binary is accessible after installation
- Fix migrations directory path from 'migrations' to './migrations' in all test jobs
- Apply changes to api-server-test, charging-engine-test, and packet-gateway-test jobs
- Add std::process::Command import for running external commands
- Implement run_migrations function to execute goose postgres migrations
- Call run_migrations in RatingPlansRepo::connect before returning repository instance
- Update function comments to reflect automatic migration execution on startup
- Add error handling and output capture for migration failures
- Add tracing log for successful migration
- Upgrade PostgreSQL service from version 15 to 18 in api-server-test job
- Add PostgreSQL 18 service configuration to charging-engine-test and packet-gateway-test jobs
- Configure database health checks with pg_isready command for all Postgres services
- Add goose installation step to charging-engine-test and packet-gateway-test jobs
- Add database migration execution step before running tests in Rust components
- Replace custom Go migration runner (go run cmd/migrate/main.go) with goose postgres command in db-migrate target
- Add db-migrate-down target for rolling back migrations
- Add db-migrate-status target to show current migration status
- Add db-create-migration target for creating new migration files with NAME parameter validation
- Add db-migrate-rust target for Rust service migrations (uses same
- Add os/exec import for running external commands
- Implement runMigrations function to execute goose postgres migrations
- Call runMigrations in NewPostgresProfileStore before returning store instance
- Update function comment to reflect automatic migration execution on startup
- Add error handling and output capture for migration failures
- Remove AutoMigrate call for profileRow from NewPostgresProfileStore
- Update function comment to reference migration system (make db-migrate) instead of inline schema creation
- Add note that esim_profiles table is created by migration system
- Remove error handling for AutoMigrate execution
- Remove CREATE TABLE IF NOT EXISTS query for rating_plans table from connect() method
- Update comment to reference migration system (make db-migrate) instead of inline schema creation
- Add note that rating_plans table is created by 000001_init_rating_plans.up.sql migration
- Remove sqlx::query execution and error mapping for schema creation
- Keep database connection pool and circuit breaker initialization
- Install goose CLI tool using go install in CI and deploy workflows
- Replace custom migration command (go run cmd/migrate/main.go --action=up) with goose postgres command
- Remove working-directory specification from CI migration step
- Update commented kubectl exec commands in deploy workflow to use goose instead of custom runner
- Simplify migration execution by using standardized goose tool across all environments
- Add migration execution step in CI workflow before linter runs
- Configure DATABASE_URL environment variable for test database connection
- Add placeholder migration steps in deploy workflow for staging and production environments
- Include commented kubectl exec commands for future implementation in Kubernetes deployments
- Add maintainer and description labels to all build and runtime stages
- Change builder WORKDIR from /app to /build for clarity
- Install build dependencies with --no-cache/--no-install-recommends flags
- Add build optimizations: -ldflags="-w -s -extldflags '-static'" and -trimpath for Go, strip for Rust
- Implement dependency caching strategy for Rust builds (dummy app pattern)
- Pin Alpine base
- Change all docker/build-push-action contexts from ./apps/<component> to root directory (.)
- Add explicit file parameter pointing to ./deployments/docker/<component>.Dockerfile for all components
- Update build configurations for api-server, carrier-connector, charging-engine, packet-gateway, CLI, and web-dashboard
- Change all docker/build-push-action contexts from ./apps/<component> to root directory (.)
- Add explicit file parameter pointing to ./deployments/docker/<component>.Dockerfile for all components
- Update build configurations for api-server, carrier-connector, charging-engine, packet-gateway, CLI, and web-dashboard
- Create multi-stage Dockerfile for CLI application
- Use golang:1.26-alpine as builder image
- Copy go.mod and go.sum for dependency caching
- Build static binary with CGO_ENABLED=0
- Use alpine:latest as runtime image with ca-certificates
- Set /root/ as working directory and ./cli as default command
- Remove duplicate CreditManager and RatingPlansRepo imports from engine.rs
- Add circuit_breaker module to main.rs imports
- Comment out rate_limit module in lib.rs
- Update rate_limit.rs imports to use GovernorConfig and GovernorLayer directly
- Change GovernorConfigBuilder to GovernorConfig::builder() and update config initialization
- Update GovernorLayer::new() to take config reference
- Add ebpf feature flag to Cargo.toml with default disabled
- Make aya and aya-log dependencies optional and tied to ebpf feature
- Add #[cfg(feature = "ebpf")] conditional compilation directives for EbpfManager imports and usage
- Add feature-gated readiness check logic (ebpf_ok only checked when feature enabled)
- Add feature-gated metrics_handler implementations (full version with eBPF, stub version without)
- Change
- Change api-server build path from default to ./cmd/ directory
- Change carrier-connector build path from default to current directory
- Change CLI build path from default to current directory
- Update build-go echo color from CYAN to GREEN for consistency
- Remove redundant mkdir -p dist from build-go target
- Change eBPF build from required to optional with graceful fallback
- Add match statement to handle build success, failure, and missing cargo-build-bpf cases
- Add cargo warnings when eBPF build fails or cargo-build-bpf is not installed
- Remove panic on build failure to allow compilation without eBPF toolchain
- Remove thiserror 2.0.18 dependency from Cargo.toml
- Comment out CircuitBreakerError import and all circuit breaker execute calls in credit_management.rs
- Comment out redis_circuit_breaker and postgres_circuit_breaker fields in ChargingEngine struct
- Comment out circuit breaker initialization code in engine.rs
- Comment out circuit_breaker field and initialization in RatingPlansRepo
- Remove circuit breaker
- Change create_rate_limiter return type from InMemoryState/DefaultClock to StateInformationMiddleware with QuantaClock
- Add StateInformationMiddleware import from governor::middleware
- Add Arc import from std::sync
- Remove explicit key_extractor field from GovernorLayer initialization
- Upgrade tower_governor from 0.4.2 to 0.8.0 and add governor 0.10.4 dependency
- Add thiserror 2.0.18 for error handling
- Update dev dependencies: tokio-test to 0.4.5, mockall to 0.14.0, serial_test to 3.4.0
- Add Copy trait to CircuitState enum
- Add Clone derive to CircuitBreaker struct
- Add error trait bounds to CircuitBreakerError generic type
- Update create_rate_limiter return type to use explicit generic parameters
- Fix Go
nutcas3
NutcaseandGitHub