Module k8s_optimize 

Kubernetes Resource Optimization Analyzer

A native Rust analyzer for detecting over-provisioned and under-provisioned Kubernetes workloads. Helps reduce cloud costs by right-sizing resource requests and limits.

Features

Phase 1: Static Analysis

• Static analysis of Kubernetes manifests (no cluster access required)
• Terraform HCL support - Parse kubernetes_* provider resources
• Pattern-based detection of over/under-provisioning
• Workload type classification for smarter recommendations
• Support for Deployments, StatefulSets, DaemonSets, Jobs, CronJobs
• Helm chart and Kustomize directory support
• Multiple output formats (table, JSON)

Phase 2: Live Cluster Analysis

• Kubernetes API integration - Connect to real clusters via kubeconfig
• metrics-server support - Real-time CPU/memory usage data
• Prometheus integration - Historical metrics (P50, P95, P99, max)
• Data-driven recommendations based on actual usage
• Waste percentage calculations with confidence levels

Example

use syncable_cli::analyzer::k8s_optimize::{lint, K8sOptimizeConfig, OptimizationResult};
use std::path::Path;

// Static analysis (no cluster needed)
let config = K8sOptimizeConfig::default();
let result = lint(Path::new("./k8s/"), &config);

// Or using the backward-compatible analyze() function:
let result = analyze(Path::new("./k8s/"), &config);

// Live cluster analysis (requires kubeconfig)
use syncable_cli::analyzer::k8s_optimize::live_analyzer::{LiveAnalyzer, LiveAnalyzerConfig};
let live_config = LiveAnalyzerConfig::default();
let analyzer = LiveAnalyzer::new(live_config).await?;
let live_result = analyzer.analyze().await?;

Optimization Rules

The analyzer checks for these common issues (K8S-OPT-001 through K8S-OPT-010):

Over-Provisioning Detection

• K8S-OPT-005: CPU request > 1 core for non-batch workload
• K8S-OPT-006: Memory request > 2Gi for non-database workload
• K8S-OPT-007: Excessive CPU limit-to-request ratio (> 10x)
• K8S-OPT-008: Excessive memory limit-to-request ratio (> 4x)

Under-Provisioning Detection

• K8S-OPT-001: No CPU request defined
• K8S-OPT-002: No memory request defined
• K8S-OPT-003: No CPU limit defined
• K8S-OPT-004: No memory limit defined

Best Practices

• K8S-OPT-009: Requests equal to limits (no bursting allowed)
• K8S-OPT-010: Unbalanced resource allocation for workload type

Re-exports

pub use config::K8sOptimizeConfig;

pub use types::AnalysisMetadata;

pub use types::AnalysisMode;

pub use types::ChartValidation;

pub use types::CloudProvider;

pub use types::CostBreakdown;

pub use types::CostEstimation;

pub use types::CostSavings;

pub use types::FixApplicationResult;

pub use types::FixImpact;

pub use types::FixResourceValues;

pub use types::FixRisk;

pub use types::FixSource;

pub use types::FixStatus;

pub use types::HelmIssue;

pub use types::HelmValidationReport;

pub use types::HelmValidationSummary;

pub use types::LiveClusterSummary;

pub use types::LiveFix;

pub use types::OptimizationIssue;

pub use types::OptimizationResult;

pub use types::OptimizationSummary;

pub use types::PreciseFix;

pub use types::ResourceOptimizationReport;

pub use types::ResourceOptimizationSummary;

pub use types::ResourceRecommendation;

pub use types::ResourceSpec;

pub use types::ResourceUsage;

pub use types::ResourceWarning;

pub use types::RuleCode;

pub use types::SecurityFinding;

pub use types::SecurityReport;

pub use types::SecuritySummary;

pub use types::Severity;

pub use types::TrendAnalysis;

pub use types::TrendDirection;

pub use types::UnifiedMetadata;

pub use types::UnifiedReport;

pub use types::UnifiedSummary;

pub use types::WasteMetrics;

pub use types::WorkloadCost;

pub use types::WorkloadTrend;

pub use types::WorkloadType;

pub use formatter::OutputFormat;

pub use formatter::format_result;

pub use formatter::format_result_to_string;

pub use static_analyzer::analyze as lint;

pub use static_analyzer::analyze_content as lint_content;

pub use static_analyzer::analyze_file as lint_file;

pub use static_analyzer::analyze;

pub use static_analyzer::analyze_content;

pub use static_analyzer::analyze_file;

pub use parser::TerraformContainer;

pub use parser::TerraformK8sResource;

pub use parser::TfResourceSpec;

pub use parser::bytes_to_memory_string;

pub use parser::cpu_limit_to_request_ratio;

pub use parser::detect_workload_type;

pub use parser::extract_container_image;

pub use parser::extract_container_name;

pub use parser::extract_resources;

pub use parser::memory_limit_to_request_ratio;

pub use parser::millicores_to_cpu_string;

pub use parser::parse_cpu_to_millicores;

pub use parser::parse_memory_to_bytes;

pub use parser::parse_terraform_k8s_resources;

pub use rules::ContainerContext;

pub use rules::OptimizationRule;

pub use rules::RuleContext;

pub use rules::all_rules;

pub use rules::codes as rule_codes;

pub use rules::generate_recommendations;

pub use rules::rule_description;

pub use pragma::IGNORE_ANNOTATION_PREFIX;

pub use pragma::extract_annotations;

pub use pragma::get_ignore_reason;

pub use pragma::get_ignored_rules;

pub use pragma::should_ignore_rule;

pub use live_analyzer::DataSource;

pub use live_analyzer::LiveAnalysisResult;

pub use live_analyzer::LiveAnalyzer;

pub use live_analyzer::LiveAnalyzerConfig;

pub use live_analyzer::LiveRecommendation;

pub use metrics_client::MetricsClient;

pub use metrics_client::PodMetrics;

pub use metrics_client::PodResources;

pub use metrics_client::ResourceComparison;

pub use prometheus_client::ContainerHistory;

pub use prometheus_client::HistoricalRecommendation;

pub use prometheus_client::PrometheusAuth;

pub use prometheus_client::PrometheusClient;

pub use cost_calculator::calculate_from_live;

pub use cost_calculator::calculate_from_static;

pub use fix_applicator::apply_fixes;

pub use fix_applicator::locate_resources_from_static;

pub use fix_applicator::locate_resources_in_file;

pub use trend_analyzer::analyze_trends_from_live;

pub use trend_analyzer::analyze_trends_static;

Modules

config

Configuration for the optimizer. Configuration for Kubernetes resource optimization analysis.

cost_calculator

Cost calculation and estimation. Cost Calculator for Kubernetes Resource Waste

fix_applicator

Fix application to manifest files. Precise Fix Locator and Applicator

formatter

Output formatting (table, JSON, YAML). Output formatting for optimization results.

live_analyzer

Live cluster analyzer. Live Cluster Analyzer for Kubernetes resource optimization.

metrics_client

Kubernetes metrics-server client. Kubernetes Metrics Client for live cluster resource usage.

parser

Parsing utilities (YAML, Terraform, Helm). Parsing utilities for Kubernetes resource analysis.

pragma

Annotation-based rule ignoring (pragma). Annotation-based rule ignoring for k8s-optimize.

prometheus_client

Prometheus client for historical metrics. Prometheus Client for historical Kubernetes metrics.

recommender

Recommendation generation (now in rules/). Resource recommendation generation.

rules

Individual optimization rules (K8S-OPT-001 through K8S-OPT-010). Individual optimization rules for Kubernetes resources.

static_analyzer

Static analysis of Kubernetes manifests. Static analysis of Kubernetes manifests for resource optimization.

terraform_parser

Terraform parser (now in parser/terraform.rs, re-exported for compatibility). Terraform HCL parser for Kubernetes resources.

trend_analyzer

Trend analysis. Trend Analyzer for Kubernetes Resource Waste

types

Core data types. Core types for Kubernetes resource optimization analysis.