- Replace hardcoded i32 enum matches with ProtoEnum::try_from() in 8 command files (cluster, config, data, model, risk, service, train, tune) - Replace hardcoded i32 literals with enum variants (EmergencyStopType, ExportFormat, TrainingMode) - Add JSON-RPC 2.0 PARSE_ERROR (-32700) for malformed JSON input - Validate jsonrpc:"2.0" version on incoming MCP requests - Change unreachable execute_tool catch-all from Ok to Err - Fix TUI teardown to not mask original event_loop error - Rename config_cmd module to config (commands::config) - 77 tests pass, 0 clippy warnings Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
464 lines
15 KiB
Rust
464 lines
15 KiB
Rust
//! `fxt service` -- service operations.
|
|
|
|
use anyhow::Result;
|
|
use clap::{Parser, Subcommand};
|
|
use colored::Colorize;
|
|
use serde::Serialize;
|
|
|
|
use crate::grpc::FoxhuntClient;
|
|
use crate::output::{self, HumanReadable, OutputFormat};
|
|
use crate::proto::monitoring;
|
|
|
|
#[derive(Parser, Debug)]
|
|
pub struct ServiceCommand {
|
|
#[command(subcommand)]
|
|
action: ServiceAction,
|
|
}
|
|
|
|
#[derive(Subcommand, Debug)]
|
|
enum ServiceAction {
|
|
/// List all services with health status
|
|
List,
|
|
/// Detailed service status
|
|
Status {
|
|
/// Service name
|
|
service: String,
|
|
},
|
|
/// Stream service logs
|
|
Logs {
|
|
/// Service name
|
|
service: String,
|
|
/// Follow log output
|
|
#[arg(long, short)]
|
|
follow: bool,
|
|
},
|
|
/// Restart a service
|
|
Restart {
|
|
/// Service name
|
|
service: String,
|
|
},
|
|
/// Deploy binary update
|
|
Deploy {
|
|
/// Service name
|
|
service: String,
|
|
},
|
|
/// Health check all services
|
|
Health,
|
|
}
|
|
|
|
// ── Result types ──────────────────────────────────────────────────────
|
|
|
|
#[derive(Serialize)]
|
|
struct ServiceListResult {
|
|
overall_health: String,
|
|
healthy: i32,
|
|
total: i32,
|
|
uptime_seconds: i64,
|
|
services: Vec<ServiceEntry>,
|
|
}
|
|
|
|
#[derive(Serialize)]
|
|
struct ServiceEntry {
|
|
name: String,
|
|
health: String,
|
|
state: String,
|
|
version: String,
|
|
uptime_seconds: i64,
|
|
error: String,
|
|
}
|
|
|
|
#[derive(Serialize)]
|
|
struct ServiceStatusResult {
|
|
name: String,
|
|
health: String,
|
|
state: String,
|
|
version: String,
|
|
uptime_seconds: i64,
|
|
error: String,
|
|
metadata: Vec<(String, String)>,
|
|
dependencies: Vec<DependencyEntry>,
|
|
}
|
|
|
|
#[derive(Serialize)]
|
|
struct DependencyEntry {
|
|
name: String,
|
|
dep_type: String,
|
|
status: String,
|
|
endpoint: String,
|
|
response_time_ms: f64,
|
|
}
|
|
|
|
#[derive(Serialize)]
|
|
struct HealthCheckResult {
|
|
overall: String,
|
|
checks: Vec<HealthEntry>,
|
|
}
|
|
|
|
#[derive(Serialize)]
|
|
struct HealthEntry {
|
|
name: String,
|
|
status: String,
|
|
message: String,
|
|
response_time_ms: f64,
|
|
}
|
|
|
|
// ── Display helpers ──────────────────────────────────────────────────
|
|
|
|
fn service_health_str(h: i32) -> &'static str {
|
|
use crate::proto::monitoring::ServiceHealth;
|
|
match ServiceHealth::try_from(h) {
|
|
Ok(ServiceHealth::Healthy) => "healthy",
|
|
Ok(ServiceHealth::Degraded) => "degraded",
|
|
Ok(ServiceHealth::Unhealthy) => "unhealthy",
|
|
Ok(ServiceHealth::Critical) => "critical",
|
|
_ => "unknown",
|
|
}
|
|
}
|
|
|
|
fn service_state_str(s: i32) -> &'static str {
|
|
use crate::proto::monitoring::ServiceState;
|
|
match ServiceState::try_from(s) {
|
|
Ok(ServiceState::Starting) => "starting",
|
|
Ok(ServiceState::Running) => "running",
|
|
Ok(ServiceState::Stopping) => "stopping",
|
|
Ok(ServiceState::Stopped) => "stopped",
|
|
Ok(ServiceState::Error) => "error",
|
|
_ => "unknown",
|
|
}
|
|
}
|
|
|
|
fn system_health_str(h: i32) -> &'static str {
|
|
use crate::proto::monitoring::SystemHealth;
|
|
match SystemHealth::try_from(h) {
|
|
Ok(SystemHealth::Healthy) => "healthy",
|
|
Ok(SystemHealth::Degraded) => "degraded",
|
|
Ok(SystemHealth::Unhealthy) => "unhealthy",
|
|
Ok(SystemHealth::Critical) => "critical",
|
|
_ => "unknown",
|
|
}
|
|
}
|
|
|
|
fn health_status_str(h: i32) -> &'static str {
|
|
use crate::proto::monitoring::HealthStatus;
|
|
match HealthStatus::try_from(h) {
|
|
Ok(HealthStatus::Healthy) => "healthy",
|
|
Ok(HealthStatus::Degraded) => "degraded",
|
|
Ok(HealthStatus::Unhealthy) => "unhealthy",
|
|
Ok(HealthStatus::Critical) => "critical",
|
|
_ => "unknown",
|
|
}
|
|
}
|
|
|
|
fn dependency_type_str(t: i32) -> &'static str {
|
|
use crate::proto::monitoring::DependencyType;
|
|
match DependencyType::try_from(t) {
|
|
Ok(DependencyType::Database) => "database",
|
|
Ok(DependencyType::MessageQueue) => "message_queue",
|
|
Ok(DependencyType::Cache) => "cache",
|
|
Ok(DependencyType::ExternalApi) => "external_api",
|
|
Ok(DependencyType::FileSystem) => "file_system",
|
|
Ok(DependencyType::Network) => "network",
|
|
_ => "unknown",
|
|
}
|
|
}
|
|
|
|
fn colorize_health(s: &str) -> String {
|
|
match s {
|
|
"healthy" => s.green().bold().to_string(),
|
|
"degraded" => s.yellow().bold().to_string(),
|
|
"unhealthy" => s.red().to_string(),
|
|
"critical" => s.red().bold().to_string(),
|
|
_ => s.dimmed().to_string(),
|
|
}
|
|
}
|
|
|
|
fn colorize_state(s: &str) -> String {
|
|
match s {
|
|
"running" => s.green().to_string(),
|
|
"starting" | "stopping" => s.yellow().to_string(),
|
|
"stopped" => s.dimmed().to_string(),
|
|
"error" => s.red().bold().to_string(),
|
|
_ => s.dimmed().to_string(),
|
|
}
|
|
}
|
|
|
|
#[allow(clippy::integer_division, clippy::modulo_arithmetic)]
|
|
fn format_uptime(secs: i64) -> String {
|
|
if secs < 60 {
|
|
return format!("{secs}s");
|
|
}
|
|
let days = secs / 86400;
|
|
let hours = (secs % 86400) / 3600;
|
|
let mins = (secs % 3600) / 60;
|
|
if days > 0 {
|
|
format!("{days}d {hours}h {mins}m")
|
|
} else if hours > 0 {
|
|
format!("{hours}h {mins}m")
|
|
} else {
|
|
format!("{mins}m")
|
|
}
|
|
}
|
|
|
|
// ── HumanReadable impls ──────────────────────────────────────────────
|
|
|
|
impl HumanReadable for ServiceListResult {
|
|
fn print_human(&self) {
|
|
println!(
|
|
"{} {} ({}/{} healthy, uptime {})\n",
|
|
"System:".bold(),
|
|
colorize_health(&self.overall_health),
|
|
self.healthy,
|
|
self.total,
|
|
format_uptime(self.uptime_seconds).dimmed(),
|
|
);
|
|
|
|
println!(
|
|
" {:<28} {:<12} {:<12} {:<10} {}",
|
|
"SERVICE".bold(),
|
|
"HEALTH".bold(),
|
|
"STATE".bold(),
|
|
"UPTIME".bold(),
|
|
"VERSION".bold(),
|
|
);
|
|
println!(" {}", "-".repeat(76));
|
|
|
|
for svc in &self.services {
|
|
let err_suffix = if svc.error.is_empty() {
|
|
String::new()
|
|
} else {
|
|
format!(" {}", svc.error.red())
|
|
};
|
|
println!(
|
|
" {:<28} {:<12} {:<12} {:<10} {}{}",
|
|
svc.name,
|
|
colorize_health(&svc.health),
|
|
colorize_state(&svc.state),
|
|
format_uptime(svc.uptime_seconds),
|
|
svc.version.dimmed(),
|
|
err_suffix,
|
|
);
|
|
}
|
|
}
|
|
}
|
|
|
|
impl HumanReadable for ServiceStatusResult {
|
|
fn print_human(&self) {
|
|
println!("{} {}", "Service:".bold(), self.name);
|
|
println!(" Health: {}", colorize_health(&self.health));
|
|
println!(" State: {}", colorize_state(&self.state));
|
|
println!(" Version: {}", self.version);
|
|
println!(" Uptime: {}", format_uptime(self.uptime_seconds));
|
|
|
|
if !self.error.is_empty() {
|
|
println!(" Error: {}", self.error.red());
|
|
}
|
|
|
|
if !self.metadata.is_empty() {
|
|
println!("\n {}:", "Metadata".bold());
|
|
for (k, v) in &self.metadata {
|
|
println!(" {k}: {v}");
|
|
}
|
|
}
|
|
|
|
if !self.dependencies.is_empty() {
|
|
println!("\n {}:", "Dependencies".bold());
|
|
for dep in &self.dependencies {
|
|
let status_col = colorize_health(&dep.status);
|
|
let ep = if dep.endpoint.is_empty() {
|
|
String::new()
|
|
} else {
|
|
format!(" ({})", dep.endpoint.dimmed())
|
|
};
|
|
println!(
|
|
" {} [{}] {} {:.1}ms{}",
|
|
dep.name, dep.dep_type, status_col, dep.response_time_ms, ep,
|
|
);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
impl HumanReadable for HealthCheckResult {
|
|
fn print_human(&self) {
|
|
println!("{} {}\n", "Overall:".bold(), colorize_health(&self.overall));
|
|
|
|
println!(
|
|
" {:<32} {:<12} {:>10} {}",
|
|
"CHECK".bold(),
|
|
"STATUS".bold(),
|
|
"LATENCY".bold(),
|
|
"MESSAGE".bold(),
|
|
);
|
|
println!(" {}", "-".repeat(72));
|
|
|
|
for c in &self.checks {
|
|
let latency_str = if c.response_time_ms > 0.0 {
|
|
format!("{:.1}ms", c.response_time_ms)
|
|
} else {
|
|
"-".to_owned()
|
|
};
|
|
println!(
|
|
" {:<32} {:<12} {:>10} {}",
|
|
c.name,
|
|
colorize_health(&c.status),
|
|
latency_str,
|
|
c.message.dimmed(),
|
|
);
|
|
}
|
|
}
|
|
}
|
|
|
|
// ── Stub result for non-gRPC commands ────────────────────────────────
|
|
|
|
#[derive(Serialize)]
|
|
struct StubMessage {
|
|
message: String,
|
|
}
|
|
|
|
impl HumanReadable for StubMessage {
|
|
fn print_human(&self) {
|
|
println!("{}", self.message);
|
|
}
|
|
}
|
|
|
|
// ── Execute ──────────────────────────────────────────────────────────
|
|
|
|
impl ServiceCommand {
|
|
pub async fn execute(&self, client: &FoxhuntClient, format: OutputFormat) -> Result<()> {
|
|
match &self.action {
|
|
ServiceAction::List => {
|
|
let resp = client
|
|
.monitoring()
|
|
.get_system_status(monitoring::GetSystemStatusRequest {
|
|
service_names: vec![],
|
|
})
|
|
.await?
|
|
.into_inner();
|
|
|
|
let overall = resp.overall_status.unwrap_or_default();
|
|
let services: Vec<ServiceEntry> = resp
|
|
.service_statuses
|
|
.iter()
|
|
.map(|s| ServiceEntry {
|
|
name: s.service_name.clone(),
|
|
health: service_health_str(s.health).to_owned(),
|
|
state: service_state_str(s.state).to_owned(),
|
|
version: s.version.clone().unwrap_or_default(),
|
|
uptime_seconds: s.uptime_seconds,
|
|
error: s.error_message.clone().unwrap_or_default(),
|
|
})
|
|
.collect();
|
|
|
|
let result = ServiceListResult {
|
|
overall_health: system_health_str(overall.overall_health).to_owned(),
|
|
healthy: overall.healthy_services,
|
|
total: overall.total_services,
|
|
uptime_seconds: overall.system_uptime_seconds,
|
|
services,
|
|
};
|
|
output::render(&result, format)?;
|
|
}
|
|
ServiceAction::Status { service } => {
|
|
let resp = client
|
|
.monitoring()
|
|
.get_system_status(monitoring::GetSystemStatusRequest {
|
|
service_names: vec![service.clone()],
|
|
})
|
|
.await?
|
|
.into_inner();
|
|
|
|
let maybe_svc = resp
|
|
.service_statuses
|
|
.into_iter()
|
|
.find(|s| s.service_name == *service);
|
|
|
|
let Some(svc) = maybe_svc else {
|
|
anyhow::bail!("service '{}' not found in system status", service);
|
|
};
|
|
|
|
let metadata: Vec<(String, String)> =
|
|
svc.metadata.into_iter().collect();
|
|
|
|
let dependencies: Vec<DependencyEntry> = svc
|
|
.dependencies
|
|
.iter()
|
|
.map(|d| DependencyEntry {
|
|
name: d.name.clone(),
|
|
dep_type: dependency_type_str(d.dependency_type).to_owned(),
|
|
status: health_status_str(d.status).to_owned(),
|
|
endpoint: d.endpoint.clone().unwrap_or_default(),
|
|
response_time_ms: d.response_time_ms.unwrap_or_default(),
|
|
})
|
|
.collect();
|
|
|
|
let result = ServiceStatusResult {
|
|
name: svc.service_name,
|
|
health: service_health_str(svc.health).to_owned(),
|
|
state: service_state_str(svc.state).to_owned(),
|
|
version: svc.version.unwrap_or_default(),
|
|
uptime_seconds: svc.uptime_seconds,
|
|
error: svc.error_message.unwrap_or_default(),
|
|
metadata,
|
|
dependencies,
|
|
};
|
|
output::render(&result, format)?;
|
|
}
|
|
ServiceAction::Health => {
|
|
let resp = client
|
|
.monitoring()
|
|
.get_health_check(monitoring::GetHealthCheckRequest {
|
|
service_name: None,
|
|
})
|
|
.await?
|
|
.into_inner();
|
|
|
|
let checks: Vec<HealthEntry> = resp
|
|
.health_checks
|
|
.iter()
|
|
.map(|c| HealthEntry {
|
|
name: c.check_name.clone(),
|
|
status: health_status_str(c.status).to_owned(),
|
|
message: c.message.clone().unwrap_or_default(),
|
|
response_time_ms: c.response_time_ms.unwrap_or_default(),
|
|
})
|
|
.collect();
|
|
|
|
let result = HealthCheckResult {
|
|
overall: health_status_str(resp.health_status).to_owned(),
|
|
checks,
|
|
};
|
|
output::render(&result, format)?;
|
|
}
|
|
ServiceAction::Logs { service, follow: _ } => {
|
|
let result = StubMessage {
|
|
message: format!(
|
|
"Service logs are available via kubectl:\n \
|
|
kubectl logs -f deploy/{service} -n foxhunt"
|
|
),
|
|
};
|
|
output::render(&result, format)?;
|
|
}
|
|
ServiceAction::Restart { service } => {
|
|
let result = StubMessage {
|
|
message: format!(
|
|
"Service restart requires kubectl access:\n \
|
|
kubectl rollout restart deploy/{service} -n foxhunt"
|
|
),
|
|
};
|
|
output::render(&result, format)?;
|
|
}
|
|
ServiceAction::Deploy { service } => {
|
|
let result = StubMessage {
|
|
message: format!(
|
|
"Use pod-writer-deploy for deployment:\n \
|
|
kubectl apply -f infra/k8s/services/{service}.yaml && \
|
|
kubectl rollout restart deploy/{service} -n foxhunt"
|
|
),
|
|
};
|
|
output::render(&result, format)?;
|
|
}
|
|
}
|
|
Ok(())
|
|
}
|
|
}
|