Merge branch 'main' of github.com:rustfs/s3-rustfs into feature/observability-metrics

# Conflicts:
#	Cargo.toml
#	crates/obs/examples/config.toml
#	crates/obs/src/telemetry.rs
This commit is contained in:
houseme
2025-05-12 15:21:30 +08:00
123 changed files with 3434 additions and 1688 deletions
+17
View File
@@ -0,0 +1,17 @@
[package]
name = "rustfs-config"
edition.workspace = true
license.workspace = true
repository.workspace = true
rust-version.workspace = true
version.workspace = true
[dependencies]
config = { workspace = true }
const-str = { workspace = true }
serde = { workspace = true }
serde_json = { workspace = true }
[lints]
workspace = true
+23
View File
@@ -0,0 +1,23 @@
use crate::event::config::NotifierConfig;
use crate::ObservabilityConfig;
/// RustFs configuration
pub struct RustFsConfig {
pub observability: ObservabilityConfig,
pub event: NotifierConfig,
}
impl RustFsConfig {
pub fn new() -> Self {
Self {
observability: ObservabilityConfig::new(),
event: NotifierConfig::new(),
}
}
}
impl Default for RustFsConfig {
fn default() -> Self {
Self::new()
}
}
+91
View File
@@ -0,0 +1,91 @@
use const_str::concat;
/// Application name
/// Default value: RustFs
/// Environment variable: RUSTFS_APP_NAME
pub const APP_NAME: &str = "RustFs";
/// Application version
/// Default value: 1.0.0
/// Environment variable: RUSTFS_VERSION
pub const VERSION: &str = "0.0.1";
/// Default configuration logger level
/// Default value: info
/// Environment variable: RUSTFS_LOG_LEVEL
pub const DEFAULT_LOG_LEVEL: &str = "info";
/// Default configuration use stdout
/// Default value: true
pub const USE_STDOUT: bool = true;
/// Default configuration sample ratio
/// Default value: 1.0
pub const SAMPLE_RATIO: f64 = 1.0;
/// Default configuration meter interval
/// Default value: 30
pub const METER_INTERVAL: u64 = 30;
/// Default configuration service version
/// Default value: 0.0.1
pub const SERVICE_VERSION: &str = "0.0.1";
/// Default configuration environment
/// Default value: production
pub const ENVIRONMENT: &str = "production";
/// maximum number of connections
/// This is the maximum number of connections that the server will accept.
/// This is used to limit the number of connections to the server.
pub const MAX_CONNECTIONS: usize = 100;
/// timeout for connections
/// This is the timeout for connections to the server.
/// This is used to limit the time that a connection can be open.
pub const DEFAULT_TIMEOUT_MS: u64 = 3000;
/// Default Access Key
/// Default value: rustfsadmin
/// Environment variable: RUSTFS_ACCESS_KEY
/// Command line argument: --access-key
/// Example: RUSTFS_ACCESS_KEY=rustfsadmin
/// Example: --access-key rustfsadmin
pub const DEFAULT_ACCESS_KEY: &str = "rustfsadmin";
/// Default Secret Key
/// Default value: rustfsadmin
/// Environment variable: RUSTFS_SECRET_KEY
/// Command line argument: --secret-key
/// Example: RUSTFS_SECRET_KEY=rustfsadmin
/// Example: --secret-key rustfsadmin
pub const DEFAULT_SECRET_KEY: &str = "rustfsadmin";
/// Default configuration file for observability
/// Default value: config/obs.toml
/// Environment variable: RUSTFS_OBS_CONFIG
/// Command line argument: --obs-config
/// Example: RUSTFS_OBS_CONFIG=config/obs.toml
/// Example: --obs-config config/obs.toml
/// Example: --obs-config /etc/rustfs/obs.toml
pub const DEFAULT_OBS_CONFIG: &str = "./deploy/config/obs.toml";
/// Default TLS key for rustfs
/// This is the default key for TLS.
pub const RUSTFS_TLS_KEY: &str = "rustfs_key.pem";
/// Default TLS cert for rustfs
/// This is the default cert for TLS.
pub const RUSTFS_TLS_CERT: &str = "rustfs_cert.pem";
/// Default port for rustfs
/// This is the default port for rustfs.
/// This is used to bind the server to a specific port.
pub const DEFAULT_PORT: u16 = 9000;
/// Default address for rustfs
/// This is the default address for rustfs.
pub const DEFAULT_ADDRESS: &str = concat!(":", DEFAULT_PORT);
/// Default port for rustfs console
/// This is the default port for rustfs console.
pub const DEFAULT_CONSOLE_PORT: u16 = 9002;
/// Default address for rustfs console
/// This is the default address for rustfs console.
pub const DEFAULT_CONSOLE_ADDRESS: &str = concat!(":", DEFAULT_CONSOLE_PORT);
+1
View File
@@ -0,0 +1 @@
pub(crate) mod app;
+27
View File
@@ -0,0 +1,27 @@
use crate::event::kafka::KafkaAdapter;
use crate::event::mqtt::MqttAdapter;
use crate::event::webhook::WebhookAdapter;
use serde::{Deserialize, Serialize};
/// Configuration for the notification system.
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(tag = "type")]
pub enum AdapterConfig {
Webhook(WebhookAdapter),
Kafka(KafkaAdapter),
Mqtt(MqttAdapter),
}
impl AdapterConfig {
/// create a new configuration with default values
pub fn new() -> Self {
Self::Webhook(WebhookAdapter::new())
}
}
impl Default for AdapterConfig {
/// create a new configuration with default values
fn default() -> Self {
Self::new()
}
}
+43
View File
@@ -0,0 +1,43 @@
use crate::event::adapters::AdapterConfig;
use serde::{Deserialize, Serialize};
use std::env;
#[allow(dead_code)]
const DEFAULT_CONFIG_FILE: &str = "event";
/// Configuration for the notification system.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct NotifierConfig {
#[serde(default = "default_store_path")]
pub store_path: String,
#[serde(default = "default_channel_capacity")]
pub channel_capacity: usize,
pub adapters: Vec<AdapterConfig>,
}
impl Default for NotifierConfig {
fn default() -> Self {
Self::new()
}
}
impl NotifierConfig {
/// create a new configuration with default values
pub fn new() -> Self {
Self {
store_path: default_store_path(),
channel_capacity: default_channel_capacity(),
adapters: vec![AdapterConfig::new()],
}
}
}
/// Provide temporary directories as default storage paths
fn default_store_path() -> String {
env::temp_dir().join("event-notification").to_string_lossy().to_string()
}
/// Provides the recommended default channel capacity for high concurrency systems
fn default_channel_capacity() -> usize {
10000 // Reasonable default values for high concurrency systems
}
+29
View File
@@ -0,0 +1,29 @@
use serde::{Deserialize, Serialize};
/// Configuration for the Kafka adapter.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct KafkaAdapter {
pub brokers: String,
pub topic: String,
pub max_retries: u32,
pub timeout: u64,
}
impl KafkaAdapter {
/// create a new configuration with default values
pub fn new() -> Self {
Self {
brokers: "localhost:9092".to_string(),
topic: "kafka_topic".to_string(),
max_retries: 3,
timeout: 1000,
}
}
}
impl Default for KafkaAdapter {
/// create a new configuration with default values
fn default() -> Self {
Self::new()
}
}
+5
View File
@@ -0,0 +1,5 @@
pub(crate) mod adapters;
pub(crate) mod config;
pub(crate) mod kafka;
pub(crate) mod mqtt;
pub(crate) mod webhook;
+31
View File
@@ -0,0 +1,31 @@
use serde::{Deserialize, Serialize};
/// Configuration for the MQTT adapter.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct MqttAdapter {
pub broker: String,
pub port: u16,
pub client_id: String,
pub topic: String,
pub max_retries: u32,
}
impl MqttAdapter {
/// create a new configuration with default values
pub fn new() -> Self {
Self {
broker: "localhost".to_string(),
port: 1883,
client_id: "mqtt_client".to_string(),
topic: "mqtt_topic".to_string(),
max_retries: 3,
}
}
}
impl Default for MqttAdapter {
/// create a new configuration with default values
fn default() -> Self {
Self::new()
}
}
+51
View File
@@ -0,0 +1,51 @@
use serde::{Deserialize, Serialize};
use std::collections::HashMap;
/// Configuration for the notification system.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct WebhookAdapter {
pub endpoint: String,
pub auth_token: Option<String>,
pub custom_headers: Option<HashMap<String, String>>,
pub max_retries: u32,
pub timeout: u64,
}
impl WebhookAdapter {
/// verify that the configuration is valid
pub fn validate(&self) -> Result<(), String> {
// verify that endpoint cannot be empty
if self.endpoint.trim().is_empty() {
return Err("Webhook endpoint cannot be empty".to_string());
}
// verification timeout must be reasonable
if self.timeout == 0 {
return Err("Webhook timeout must be greater than 0".to_string());
}
// Verify that the maximum number of retry is reasonable
if self.max_retries > 10 {
return Err("Maximum retry count cannot exceed 10".to_string());
}
Ok(())
}
/// Get the default configuration
pub fn new() -> Self {
Self {
endpoint: "".to_string(),
auth_token: None,
custom_headers: Some(HashMap::new()),
max_retries: 3,
timeout: 1000,
}
}
}
impl Default for WebhookAdapter {
fn default() -> Self {
Self::new()
}
}
+11
View File
@@ -0,0 +1,11 @@
use crate::observability::config::ObservabilityConfig;
mod config;
mod constants;
mod event;
mod observability;
pub use config::RustFsConfig;
pub use constants::app::*;
pub use event::config::NotifierConfig;
+28
View File
@@ -0,0 +1,28 @@
use crate::observability::logger::LoggerConfig;
use crate::observability::otel::OtelConfig;
use crate::observability::sink::SinkConfig;
use serde::{Deserialize, Serialize};
/// Observability configuration
#[derive(Debug, Deserialize, Serialize, Clone)]
pub struct ObservabilityConfig {
pub otel: OtelConfig,
pub sinks: Vec<SinkConfig>,
pub logger: Option<LoggerConfig>,
}
impl ObservabilityConfig {
pub fn new() -> Self {
Self {
otel: OtelConfig::new(),
sinks: vec![SinkConfig::new()],
logger: Some(LoggerConfig::new()),
}
}
}
impl Default for ObservabilityConfig {
fn default() -> Self {
Self::new()
}
}
+59
View File
@@ -0,0 +1,59 @@
use serde::{Deserialize, Serialize};
use std::env;
/// File sink configuration
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct FileSink {
pub path: String,
#[serde(default = "default_buffer_size")]
pub buffer_size: Option<usize>,
#[serde(default = "default_flush_interval_ms")]
pub flush_interval_ms: Option<u64>,
#[serde(default = "default_flush_threshold")]
pub flush_threshold: Option<usize>,
}
impl FileSink {
pub fn new() -> Self {
Self {
path: env::var("RUSTFS_SINKS_FILE_PATH")
.ok()
.filter(|s| !s.trim().is_empty())
.unwrap_or_else(default_path),
buffer_size: default_buffer_size(),
flush_interval_ms: default_flush_interval_ms(),
flush_threshold: default_flush_threshold(),
}
}
}
impl Default for FileSink {
fn default() -> Self {
Self::new()
}
}
fn default_buffer_size() -> Option<usize> {
Some(8192)
}
fn default_flush_interval_ms() -> Option<u64> {
Some(1000)
}
fn default_flush_threshold() -> Option<usize> {
Some(100)
}
fn default_path() -> String {
let temp_dir = env::temp_dir().join("rustfs");
if let Err(e) = std::fs::create_dir_all(&temp_dir) {
eprintln!("Failed to create log directory: {}", e);
return "rustfs/rustfs.log".to_string();
}
temp_dir
.join("rustfs.log")
.to_str()
.unwrap_or("rustfs/rustfs.log")
.to_string()
}
+36
View File
@@ -0,0 +1,36 @@
use serde::{Deserialize, Serialize};
/// Kafka sink configuration
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct KafkaSink {
pub brokers: String,
pub topic: String,
#[serde(default = "default_batch_size")]
pub batch_size: Option<usize>,
#[serde(default = "default_batch_timeout_ms")]
pub batch_timeout_ms: Option<u64>,
}
impl KafkaSink {
pub fn new() -> Self {
Self {
brokers: "localhost:9092".to_string(),
topic: "rustfs".to_string(),
batch_size: default_batch_size(),
batch_timeout_ms: default_batch_timeout_ms(),
}
}
}
impl Default for KafkaSink {
fn default() -> Self {
Self::new()
}
}
fn default_batch_size() -> Option<usize> {
Some(100)
}
fn default_batch_timeout_ms() -> Option<u64> {
Some(1000)
}
+21
View File
@@ -0,0 +1,21 @@
use serde::{Deserialize, Serialize};
/// Logger configuration
#[derive(Debug, Deserialize, Serialize, Clone)]
pub struct LoggerConfig {
pub queue_capacity: Option<usize>,
}
impl LoggerConfig {
pub fn new() -> Self {
Self {
queue_capacity: Some(10000),
}
}
}
impl Default for LoggerConfig {
fn default() -> Self {
Self::new()
}
}
+7
View File
@@ -0,0 +1,7 @@
pub(crate) mod config;
pub(crate) mod file;
pub(crate) mod kafka;
pub(crate) mod logger;
pub(crate) mod otel;
pub(crate) mod sink;
pub(crate) mod webhook;
+69
View File
@@ -0,0 +1,69 @@
use crate::constants::app::{ENVIRONMENT, METER_INTERVAL, SAMPLE_RATIO, SERVICE_VERSION, USE_STDOUT};
use crate::{APP_NAME, DEFAULT_LOG_LEVEL};
use serde::{Deserialize, Serialize};
use std::env;
/// OpenTelemetry configuration
#[derive(Debug, Deserialize, Serialize, Clone)]
pub struct OtelConfig {
pub endpoint: String, // Endpoint for metric collection
pub use_stdout: Option<bool>, // Output to stdout
pub sample_ratio: Option<f64>, // Trace sampling ratio
pub meter_interval: Option<u64>, // Metric collection interval
pub service_name: Option<String>, // Service name
pub service_version: Option<String>, // Service version
pub environment: Option<String>, // Environment
pub logger_level: Option<String>, // Logger level
pub local_logging_enabled: Option<bool>, // Local logging enabled
}
impl OtelConfig {
pub fn new() -> Self {
extract_otel_config_from_env()
}
}
impl Default for OtelConfig {
fn default() -> Self {
Self::new()
}
}
// Helper function: Extract observable configuration from environment variables
fn extract_otel_config_from_env() -> OtelConfig {
OtelConfig {
endpoint: env::var("RUSTFS_OBSERVABILITY_ENDPOINT").unwrap_or_else(|_| "".to_string()),
use_stdout: env::var("RUSTFS_OBSERVABILITY_USE_STDOUT")
.ok()
.and_then(|v| v.parse().ok())
.or(Some(USE_STDOUT)),
sample_ratio: env::var("RUSTFS_OBSERVABILITY_SAMPLE_RATIO")
.ok()
.and_then(|v| v.parse().ok())
.or(Some(SAMPLE_RATIO)),
meter_interval: env::var("RUSTFS_OBSERVABILITY_METER_INTERVAL")
.ok()
.and_then(|v| v.parse().ok())
.or(Some(METER_INTERVAL)),
service_name: env::var("RUSTFS_OBSERVABILITY_SERVICE_NAME")
.ok()
.and_then(|v| v.parse().ok())
.or(Some(APP_NAME.to_string())),
service_version: env::var("RUSTFS_OBSERVABILITY_SERVICE_VERSION")
.ok()
.and_then(|v| v.parse().ok())
.or(Some(SERVICE_VERSION.to_string())),
environment: env::var("RUSTFS_OBSERVABILITY_ENVIRONMENT")
.ok()
.and_then(|v| v.parse().ok())
.or(Some(ENVIRONMENT.to_string())),
logger_level: env::var("RUSTFS_OBSERVABILITY_LOGGER_LEVEL")
.ok()
.and_then(|v| v.parse().ok())
.or(Some(DEFAULT_LOG_LEVEL.to_string())),
local_logging_enabled: env::var("RUSTFS_OBSERVABILITY_LOCAL_LOGGING_ENABLED")
.ok()
.and_then(|v| v.parse().ok())
.or(Some(false)),
}
}
+25
View File
@@ -0,0 +1,25 @@
use crate::observability::file::FileSink;
use crate::observability::kafka::KafkaSink;
use crate::observability::webhook::WebhookSink;
use serde::{Deserialize, Serialize};
/// Sink configuration
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(tag = "type")]
pub enum SinkConfig {
Kafka(KafkaSink),
Webhook(WebhookSink),
File(FileSink),
}
impl SinkConfig {
pub fn new() -> Self {
Self::File(FileSink::new())
}
}
impl Default for SinkConfig {
fn default() -> Self {
Self::new()
}
}
@@ -0,0 +1,39 @@
use serde::{Deserialize, Serialize};
use std::collections::HashMap;
/// Webhook sink configuration
#[derive(Debug, Deserialize, Serialize, Clone)]
pub struct WebhookSink {
pub endpoint: String,
pub auth_token: String,
pub headers: Option<HashMap<String, String>>,
#[serde(default = "default_max_retries")]
pub max_retries: Option<usize>,
#[serde(default = "default_retry_delay_ms")]
pub retry_delay_ms: Option<u64>,
}
impl WebhookSink {
pub fn new() -> Self {
Self {
endpoint: "".to_string(),
auth_token: "".to_string(),
headers: Some(HashMap::new()),
max_retries: default_max_retries(),
retry_delay_ms: default_retry_delay_ms(),
}
}
}
impl Default for WebhookSink {
fn default() -> Self {
Self::new()
}
}
fn default_max_retries() -> Option<usize> {
Some(3)
}
fn default_retry_delay_ms() -> Option<u64> {
Some(100)
}
+5 -2
View File
@@ -9,13 +9,12 @@ version.workspace = true
[features]
default = ["webhook"]
webhook = ["dep:reqwest"]
kafka = ["rdkafka"]
mqtt = ["rumqttc"]
kafka = ["dep:rdkafka"]
[dependencies]
async-trait = { workspace = true }
config = { workspace = true }
rdkafka = { workspace = true, features = ["tokio"], optional = true }
reqwest = { workspace = true, optional = true }
rumqttc = { workspace = true, optional = true }
serde = { workspace = true }
@@ -29,12 +28,16 @@ tokio = { workspace = true, features = ["sync", "net", "macros", "signal", "rt-m
tokio-util = { workspace = true }
uuid = { workspace = true, features = ["v4", "serde"] }
# Only enable kafka features and related dependencies on Linux
[target.'cfg(target_os = "linux")'.dependencies]
rdkafka = { workspace = true, features = ["tokio"], optional = true }
[dev-dependencies]
tokio = { workspace = true, features = ["test-util"] }
tracing-subscriber = { workspace = true }
http = { workspace = true }
axum = { workspace = true }
dotenvy = "0.15.7"
[lints]
workspace = true
@@ -1,28 +0,0 @@
# ===== 全局配置 =====
NOTIFIER__STORE_PATH=/var/log/event-notification
NOTIFIER__CHANNEL_CAPACITY=5000
# ===== 适配器配置(数组格式) =====
# Webhook 适配器(索引 0
NOTIFIER__ADAPTERS_0__type=Webhook
NOTIFIER__ADAPTERS_0__endpoint=http://127.0.0.1:3020/webhook
NOTIFIER__ADAPTERS_0__auth_token=your-auth-token
NOTIFIER__ADAPTERS_0__max_retries=3
NOTIFIER__ADAPTERS_0__timeout=50
NOTIFIER__ADAPTERS_0__custom_headers__x_custom_server=value
NOTIFIER__ADAPTERS_0__custom_headers__x_custom_client=value
# Kafka 适配器(索引 1
NOTIFIER__ADAPTERS_1__type=Kafka
NOTIFIER__ADAPTERS_1__brokers=localhost:9092
NOTIFIER__ADAPTERS_1__topic=notifications
NOTIFIER__ADAPTERS_1__max_retries=3
NOTIFIER__ADAPTERS_1__timeout=60
# MQTT 适配器(索引 2
NOTIFIER__ADAPTERS_2__type=Mqtt
NOTIFIER__ADAPTERS_2__broker=mqtt.example.com
NOTIFIER__ADAPTERS_2__port=1883
NOTIFIER__ADAPTERS_2__client_id=event-notifier
NOTIFIER__ADAPTERS_2__topic=events
NOTIFIER__ADAPTERS_2__max_retries=3
+28 -28
View File
@@ -1,28 +1,28 @@
# ===== global configuration =====
NOTIFIER__STORE_PATH=/var/log/event-notification
NOTIFIER__CHANNEL_CAPACITY=5000
# ===== adapter configuration array format =====
# webhook adapter index 0
NOTIFIER__ADAPTERS_0__type=Webhook
NOTIFIER__ADAPTERS_0__endpoint=http://127.0.0.1:3020/webhook
NOTIFIER__ADAPTERS_0__auth_token=your-auth-token
NOTIFIER__ADAPTERS_0__max_retries=3
NOTIFIER__ADAPTERS_0__timeout=50
NOTIFIER__ADAPTERS_0__custom_headers__x_custom_server=server-value
NOTIFIER__ADAPTERS_0__custom_headers__x_custom_client=client-value
# kafka adapter index 1
NOTIFIER__ADAPTERS_1__type=Kafka
NOTIFIER__ADAPTERS_1__brokers=localhost:9092
NOTIFIER__ADAPTERS_1__topic=notifications
NOTIFIER__ADAPTERS_1__max_retries=3
NOTIFIER__ADAPTERS_1__timeout=60
# mqtt adapter index 2
NOTIFIER__ADAPTERS_2__type=Mqtt
NOTIFIER__ADAPTERS_2__broker=mqtt.example.com
NOTIFIER__ADAPTERS_2__port=1883
NOTIFIER__ADAPTERS_2__client_id=event-notifier
NOTIFIER__ADAPTERS_2__topic=events
NOTIFIER__ADAPTERS_2__max_retries=3
## ===== global configuration =====
#NOTIFIER__STORE_PATH=/var/log/event-notification
#NOTIFIER__CHANNEL_CAPACITY=5000
#
## ===== adapter configuration array format =====
## webhook adapter index 0
#NOTIFIER__ADAPTERS_0__type=Webhook
#NOTIFIER__ADAPTERS_0__endpoint=http://127.0.0.1:3020/webhook
#NOTIFIER__ADAPTERS_0__auth_token=your-auth-token
#NOTIFIER__ADAPTERS_0__max_retries=3
#NOTIFIER__ADAPTERS_0__timeout=50
#NOTIFIER__ADAPTERS_0__custom_headers__x_custom_server=server-value
#NOTIFIER__ADAPTERS_0__custom_headers__x_custom_client=client-value
#
## kafka adapter index 1
#NOTIFIER__ADAPTERS_1__type=Kafka
#NOTIFIER__ADAPTERS_1__brokers=localhost:9092
#NOTIFIER__ADAPTERS_1__topic=notifications
#NOTIFIER__ADAPTERS_1__max_retries=3
#NOTIFIER__ADAPTERS_1__timeout=60
#
## mqtt adapter index 2
#NOTIFIER__ADAPTERS_2__type=Mqtt
#NOTIFIER__ADAPTERS_2__broker=mqtt.example.com
#NOTIFIER__ADAPTERS_2__port=1883
#NOTIFIER__ADAPTERS_2__client_id=event-notifier
#NOTIFIER__ADAPTERS_2__topic=events
#NOTIFIER__ADAPTERS_2__max_retries=3
@@ -0,0 +1,28 @@
## ===== 全局配置 =====
#NOTIFIER__STORE_PATH=/var/log/event-notification
#NOTIFIER__CHANNEL_CAPACITY=5000
#
## ===== 适配器配置(数组格式) =====
## Webhook 适配器(索引 0
#NOTIFIER__ADAPTERS_0__type=Webhook
#NOTIFIER__ADAPTERS_0__endpoint=http://127.0.0.1:3020/webhook
#NOTIFIER__ADAPTERS_0__auth_token=your-auth-token
#NOTIFIER__ADAPTERS_0__max_retries=3
#NOTIFIER__ADAPTERS_0__timeout=50
#NOTIFIER__ADAPTERS_0__custom_headers__x_custom_server=value
#NOTIFIER__ADAPTERS_0__custom_headers__x_custom_client=value
#
## Kafka 适配器(索引 1
#NOTIFIER__ADAPTERS_1__type=Kafka
#NOTIFIER__ADAPTERS_1__brokers=localhost:9092
#NOTIFIER__ADAPTERS_1__topic=notifications
#NOTIFIER__ADAPTERS_1__max_retries=3
#NOTIFIER__ADAPTERS_1__timeout=60
#
## MQTT 适配器(索引 2
#NOTIFIER__ADAPTERS_2__type=Mqtt
#NOTIFIER__ADAPTERS_2__broker=mqtt.example.com
#NOTIFIER__ADAPTERS_2__port=1883
#NOTIFIER__ADAPTERS_2__client_id=event-notifier
#NOTIFIER__ADAPTERS_2__topic=events
#NOTIFIER__ADAPTERS_2__max_retries=3
+3 -1
View File
@@ -33,7 +33,9 @@ async fn main() -> Result<(), Box<dyn error::Error>> {
// loading configuration from environment variables
let _config = NotifierConfig::event_load_config(Some("./crates/event-notifier/examples/event.toml".to_string()));
tracing::info!("event_load_config config: {:?} \n", _config);
dotenvy::dotenv()?;
let _config = NotifierConfig::event_load_config(None);
tracing::info!("event_load_config config: {:?} \n", _config);
let system = Arc::new(tokio::sync::Mutex::new(NotifierSystem::new(config.clone()).await?));
let adapters = create_adapters(&config.adapters)?;
+4 -4
View File
@@ -37,7 +37,7 @@ async fn receive_webhook(Json(payload): Json<Value>) -> StatusCode {
println!("current time:{:04}-{:02}-{:02} {:02}:{:02}:{:02}", year, month, day, hour, minute, second);
println!(
"received a webhook request time:{} content:\n {}",
seconds.to_string(),
seconds,
serde_json::to_string_pretty(&payload).unwrap()
);
StatusCode::OK
@@ -66,10 +66,10 @@ fn convert_seconds_to_date(seconds: u64) -> (u32, u32, u32, u32, u32, u32) {
// calculate month
let days_in_month = [31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31];
for m in 0..12 {
if total_seconds >= days_in_month[m] * seconds_per_day {
for m in &days_in_month {
if total_seconds >= m * seconds_per_day {
month += 1;
total_seconds -= days_in_month[m] * seconds_per_day;
total_seconds -= m * seconds_per_day;
} else {
break;
}
+3 -3
View File
@@ -4,7 +4,7 @@ use crate::Event;
use async_trait::async_trait;
use std::sync::Arc;
#[cfg(feature = "kafka")]
#[cfg(all(feature = "kafka", target_os = "linux"))]
pub(crate) mod kafka;
#[cfg(feature = "mqtt")]
pub(crate) mod mqtt;
@@ -31,7 +31,7 @@ pub fn create_adapters(configs: &[AdapterConfig]) -> Result<Vec<Arc<dyn ChannelA
webhook_config.validate().map_err(Error::ConfigError)?;
adapters.push(Arc::new(webhook::WebhookAdapter::new(webhook_config.clone())));
}
#[cfg(feature = "kafka")]
#[cfg(all(feature = "kafka", target_os = "linux"))]
AdapterConfig::Kafka(kafka_config) => {
adapters.push(Arc::new(kafka::KafkaAdapter::new(kafka_config)?));
}
@@ -43,7 +43,7 @@ pub fn create_adapters(configs: &[AdapterConfig]) -> Result<Vec<Arc<dyn ChannelA
}
#[cfg(not(feature = "webhook"))]
AdapterConfig::Webhook(_) => return Err(Error::FeatureDisabled("webhook")),
#[cfg(not(feature = "kafka"))]
#[cfg(any(not(feature = "kafka"), not(target_os = "linux")))]
AdapterConfig::Kafka(_) => return Err(Error::FeatureDisabled("kafka")),
#[cfg(not(feature = "mqtt"))]
AdapterConfig::Mqtt(_) => return Err(Error::FeatureDisabled("mqtt")),
+2
View File
@@ -7,11 +7,13 @@ use std::time::{SystemTime, UNIX_EPOCH};
use tokio::sync::mpsc;
use tokio::time::Duration;
use tokio_util::sync::CancellationToken;
use tracing::instrument;
/// Handles incoming events from the producer.
///
/// This function is responsible for receiving events from the producer and sending them to the appropriate adapters.
/// It also handles the shutdown process and saves any pending logs to the event store.
#[instrument(skip_all)]
pub async fn event_bus(
mut rx: mpsc::Receiver<Event>,
adapters: Vec<Arc<dyn ChannelAdapter>>,
+2 -11
View File
@@ -1,4 +1,4 @@
use config::{Config, Environment, File, FileFormat};
use config::{Config, File, FileFormat};
use serde::{Deserialize, Serialize};
use std::collections::HashMap;
use std::env;
@@ -138,15 +138,6 @@ impl NotifierConfig {
let app_config = Config::builder()
.add_source(File::with_name(config_dir.as_str()).format(FileFormat::Toml).required(false))
.add_source(File::with_name(config_dir.as_str()).format(FileFormat::Yaml).required(false))
.add_source(
Environment::default()
.prefix("NOTIFIER")
.prefix_separator("__")
.separator("__")
.list_separator("_")
.with_list_parse_key("adapters")
.try_parsing(true),
)
.build()
.unwrap_or_default();
match app_config.try_deserialize::<NotifierConfig>() {
@@ -162,7 +153,7 @@ impl NotifierConfig {
}
}
const DEFAULT_CONFIG_FILE: &str = "obs";
const DEFAULT_CONFIG_FILE: &str = "event";
/// Provide temporary directories as default storage paths
fn default_store_path() -> String {
+1 -1
View File
@@ -15,7 +15,7 @@ pub enum Error {
Serde(#[from] serde_json::Error),
#[error("HTTP error: {0}")]
Http(#[from] reqwest::Error),
#[cfg(feature = "kafka")]
#[cfg(all(feature = "kafka", target_os = "linux"))]
#[error("Kafka error: {0}")]
Kafka(#[from] rdkafka::error::KafkaError),
#[cfg(feature = "mqtt")]
+169
View File
@@ -15,6 +15,18 @@ pub struct Identity {
pub principal_id: String,
}
impl Identity {
/// Create a new Identity instance
pub fn new(principal_id: String) -> Self {
Self { principal_id }
}
/// Set the principal ID
pub fn set_principal_id(&mut self, principal_id: String) {
self.principal_id = principal_id;
}
}
/// A struct representing the bucket information
#[derive(Serialize, Deserialize, Clone, Debug)]
pub struct Bucket {
@@ -24,6 +36,32 @@ pub struct Bucket {
pub arn: String,
}
impl Bucket {
/// Create a new Bucket instance
pub fn new(name: String, owner_identity: Identity, arn: String) -> Self {
Self {
name,
owner_identity,
arn,
}
}
/// Set the name of the bucket
pub fn set_name(&mut self, name: String) {
self.name = name;
}
/// Set the ARN of the bucket
pub fn set_arn(&mut self, arn: String) {
self.arn = arn;
}
/// Set the owner identity of the bucket
pub fn set_owner_identity(&mut self, owner_identity: Identity) {
self.owner_identity = owner_identity;
}
}
/// A struct representing the object information
#[derive(Serialize, Deserialize, Clone, Debug)]
pub struct Object {
@@ -41,6 +79,64 @@ pub struct Object {
pub sequencer: String,
}
impl Object {
/// Create a new Object instance
pub fn new(
key: String,
size: Option<i64>,
etag: Option<String>,
content_type: Option<String>,
user_metadata: Option<HashMap<String, String>>,
version_id: Option<String>,
sequencer: String,
) -> Self {
Self {
key,
size,
etag,
content_type,
user_metadata,
version_id,
sequencer,
}
}
/// Set the key
pub fn set_key(&mut self, key: String) {
self.key = key;
}
/// Set the size
pub fn set_size(&mut self, size: Option<i64>) {
self.size = size;
}
/// Set the etag
pub fn set_etag(&mut self, etag: Option<String>) {
self.etag = etag;
}
/// Set the content type
pub fn set_content_type(&mut self, content_type: Option<String>) {
self.content_type = content_type;
}
/// Set the user metadata
pub fn set_user_metadata(&mut self, user_metadata: Option<HashMap<String, String>>) {
self.user_metadata = user_metadata;
}
/// Set the version ID
pub fn set_version_id(&mut self, version_id: Option<String>) {
self.version_id = version_id;
}
/// Set the sequencer
pub fn set_sequencer(&mut self, sequencer: String) {
self.sequencer = sequencer;
}
}
/// A struct representing the metadata of the event
#[derive(Serialize, Deserialize, Clone, Debug)]
pub struct Metadata {
@@ -52,6 +148,57 @@ pub struct Metadata {
pub object: Object,
}
impl Default for Metadata {
fn default() -> Self {
Self::new()
}
}
impl Metadata {
/// Create a new Metadata instance with default values
pub fn new() -> Self {
Self {
schema_version: "1.0".to_string(),
configuration_id: "default".to_string(),
bucket: Bucket::new(
"default".to_string(),
Identity::new("default".to_string()),
"arn:aws:s3:::default".to_string(),
),
object: Object::new("default".to_string(), None, None, None, None, None, "default".to_string()),
}
}
/// Create a new Metadata instance
pub fn create(schema_version: String, configuration_id: String, bucket: Bucket, object: Object) -> Self {
Self {
schema_version,
configuration_id,
bucket,
object,
}
}
/// Set the schema version
pub fn set_schema_version(&mut self, schema_version: String) {
self.schema_version = schema_version;
}
/// Set the configuration ID
pub fn set_configuration_id(&mut self, configuration_id: String) {
self.configuration_id = configuration_id;
}
/// Set the bucket
pub fn set_bucket(&mut self, bucket: Bucket) {
self.bucket = bucket;
}
/// Set the object
pub fn set_object(&mut self, object: Object) {
self.object = object;
}
}
/// A struct representing the source of the event
#[derive(Serialize, Deserialize, Clone, Debug)]
pub struct Source {
@@ -61,6 +208,28 @@ pub struct Source {
pub user_agent: String,
}
impl Source {
/// Create a new Source instance
pub fn new(host: String, port: String, user_agent: String) -> Self {
Self { host, port, user_agent }
}
/// Set the host
pub fn set_host(&mut self, host: String) {
self.host = host;
}
/// Set the port
pub fn set_port(&mut self, port: String) {
self.port = port;
}
/// Set the user agent
pub fn set_user_agent(&mut self, user_agent: String) {
self.user_agent = user_agent;
}
}
/// Builder for creating an Event.
///
/// This struct is used to build an Event object with various parameters.
+5 -2
View File
@@ -1,6 +1,7 @@
use crate::{create_adapters, Error, Event, NotifierConfig, NotifierSystem};
use std::sync::{atomic, Arc};
use tokio::sync::{Mutex, OnceCell};
use tracing::instrument;
static GLOBAL_SYSTEM: OnceCell<Arc<Mutex<NotifierSystem>>> = OnceCell::const_new();
static INITIALIZED: atomic::AtomicBool = atomic::AtomicBool::new(false);
@@ -113,6 +114,7 @@ pub fn is_ready() -> bool {
/// - The system is not initialized.
/// - The system is not ready.
/// - Sending the event fails.
#[instrument(fields(event))]
pub async fn send_event(event: Event) -> Result<(), Error> {
if !READY.load(atomic::Ordering::SeqCst) {
return Err(Error::custom("Notification system not ready, please wait for initialization to complete"));
@@ -124,6 +126,7 @@ pub async fn send_event(event: Event) -> Result<(), Error> {
}
/// Shuts down the notification system.
#[instrument]
pub async fn shutdown() -> Result<(), Error> {
if let Some(system) = GLOBAL_SYSTEM.get() {
tracing::info!("Shutting down notification system start");
@@ -189,7 +192,7 @@ mod tests {
let config = NotifierConfig::default();
let _ = initialize(config.clone()).await; // first initialization
let result = initialize(config).await; // second initialization
assert!(!result.is_ok(), "Initialization should succeed");
assert!(result.is_err(), "Initialization should succeed");
assert!(result.is_err(), "Re-initialization should fail");
}
@@ -211,7 +214,7 @@ mod tests {
..Default::default()
};
let result = initialize(config).await;
assert!(!result.is_err(), "Initialization with invalid config should fail");
assert!(result.is_ok(), "Initialization with invalid config should fail");
assert!(is_initialized(), "System should not be marked as initialized after failure");
assert!(is_ready(), "System should not be marked as ready after failure");
}
+2 -2
View File
@@ -8,7 +8,7 @@ mod notifier;
mod store;
pub use adapter::create_adapters;
#[cfg(feature = "kafka")]
#[cfg(all(feature = "kafka", target_os = "linux"))]
pub use adapter::kafka::KafkaAdapter;
#[cfg(feature = "mqtt")]
pub use adapter::mqtt::MqttAdapter;
@@ -16,7 +16,7 @@ pub use adapter::mqtt::MqttAdapter;
pub use adapter::webhook::WebhookAdapter;
pub use adapter::ChannelAdapter;
pub use bus::event_bus;
#[cfg(feature = "kafka")]
#[cfg(all(feature = "kafka", target_os = "linux"))]
pub use config::KafkaConfig;
#[cfg(feature = "mqtt")]
pub use config::MqttConfig;
+9 -1
View File
@@ -2,6 +2,7 @@ use crate::{event_bus, ChannelAdapter, Error, Event, EventStore, NotifierConfig}
use std::sync::Arc;
use tokio::sync::mpsc;
use tokio_util::sync::CancellationToken;
use tracing::instrument;
/// The `NotificationSystem` struct represents the notification system.
/// It manages the event bus and the adapters.
@@ -18,6 +19,7 @@ pub struct NotifierSystem {
impl NotifierSystem {
/// Creates a new `NotificationSystem` instance.
#[instrument(skip(config))]
pub async fn new(config: NotifierConfig) -> Result<Self, Error> {
let (tx, rx) = mpsc::channel::<Event>(config.channel_capacity);
let store = Arc::new(EventStore::new(&config.store_path).await?);
@@ -44,6 +46,7 @@ impl NotifierSystem {
/// Starts the notification system.
/// It initializes the event bus and the producer.
#[instrument(skip_all)]
pub async fn start(&mut self, adapters: Vec<Arc<dyn ChannelAdapter>>) -> Result<(), Error> {
if self.shutdown.is_cancelled() {
let error = Error::custom("System is shutting down");
@@ -67,6 +70,7 @@ impl NotifierSystem {
/// Sends an event to the notification system.
/// This method is used to send events to the event bus.
#[instrument(skip(self))]
pub async fn send_event(&self, event: Event) -> Result<(), Error> {
self.log(tracing::Level::DEBUG, "send_event", &format!("Sending event: {:?}", event));
if self.shutdown.is_cancelled() {
@@ -85,6 +89,7 @@ impl NotifierSystem {
/// Shuts down the notification system.
/// This method is used to cancel the event bus and producer tasks.
#[instrument(skip(self))]
pub async fn shutdown(&mut self) -> Result<(), Error> {
tracing::info!("Shutting down the notification system");
self.shutdown.cancel();
@@ -112,10 +117,13 @@ impl NotifierSystem {
self.shutdown.is_cancelled()
}
fn handle_error(&self, context: &str, error: &Error) {
#[instrument(skip(self))]
pub fn handle_error(&self, context: &str, error: &Error) {
self.log(tracing::Level::ERROR, context, &format!("{:?}", error));
// TODO Can be extended to record to files or send to monitoring systems
}
#[instrument(skip(self))]
fn log(&self, level: tracing::Level, context: &str, message: &str) {
match level {
tracing::Level::ERROR => tracing::error!("[{}] {}", context, message),
+2
View File
@@ -5,6 +5,7 @@ use std::time::{SystemTime, UNIX_EPOCH};
use tokio::fs::{create_dir_all, File, OpenOptions};
use tokio::io::{AsyncBufReadExt, AsyncWriteExt, BufReader, BufWriter};
use tokio::sync::RwLock;
use tracing::instrument;
/// `EventStore` is a struct that manages the storage of event logs.
pub struct EventStore {
@@ -21,6 +22,7 @@ impl EventStore {
})
}
#[instrument(skip(self))]
pub async fn save_logs(&self, logs: &[Log]) -> Result<(), Error> {
let _guard = self.lock.write().await;
let file_path = format!(
+6 -3
View File
@@ -13,11 +13,11 @@ workspace = true
default = ["file"]
file = []
gpu = ["dep:nvml-wrapper"]
kafka = ["dep:rdkafka"]
webhook = ["dep:reqwest"]
full = ["file", "gpu", "kafka", "webhook"]
kafka = ["dep:rdkafka"]
[dependencies]
rustfs-config = { workspace = true }
async-trait = { workspace = true }
chrono = { workspace = true }
config = { workspace = true }
@@ -29,6 +29,7 @@ opentelemetry_sdk = { workspace = true, features = ["rt-tokio"] }
opentelemetry-stdout = { workspace = true }
opentelemetry-otlp = { workspace = true, features = ["grpc-tonic", "gzip-tonic"] }
opentelemetry-semantic-conventions = { workspace = true, features = ["semconv_experimental"] }
rustfs-utils = { workspace = true, features = ["ip"] }
serde = { workspace = true }
smallvec = { workspace = true, features = ["serde"] }
tracing = { workspace = true, features = ["std", "attributes"] }
@@ -37,12 +38,14 @@ tracing-error = { workspace = true }
tracing-opentelemetry = { workspace = true }
tracing-subscriber = { workspace = true, features = ["registry", "std", "fmt", "env-filter", "tracing-log", "time", "local-time", "json"] }
tokio = { workspace = true, features = ["sync", "fs", "rt-multi-thread", "rt", "time", "macros"] }
rdkafka = { workspace = true, features = ["tokio"], optional = true }
reqwest = { workspace = true, optional = true, default-features = false }
serde_json = { workspace = true }
sysinfo = { workspace = true }
thiserror = { workspace = true }
# Only enable kafka features and related dependencies on Linux
[target.'cfg(target_os = "linux")'.dependencies]
rdkafka = { workspace = true, features = ["tokio"], optional = true }
[dev-dependencies]
+19 -18
View File
@@ -9,26 +9,27 @@ environments = "develop"
logger_level = "debug"
local_logging_enabled = true # Default is false if not specified
[sinks]
[sinks.kafka]
enabled = false
bootstrap_servers = "localhost:9092"
topic = "logs"
batch_size = 100 # Default is 100 if not specified
batch_timeout_ms = 1000 # Default is 1000ms if not specified
[sinks.webhook]
enabled = false
endpoint = "http://localhost:8080/webhook"
auth_token = ""
batch_size = 100 # Default is 3 if not specified
batch_timeout_ms = 1000 # Default is 100ms if not specified
#[[sinks]]
#type = "Kafka"
#bootstrap_servers = "localhost:9092"
#topic = "logs"
#batch_size = 100 # Default is 100 if not specified
#batch_timeout_ms = 100 # Default is 1000ms if not specified
#
#[[sinks]]
#type = "Webhook"
#endpoint = "http://localhost:8080/webhook"
#auth_token = ""
#batch_size = 100 # Default is 3 if not specified
#batch_timeout_ms = 100 # Default is 100ms if not specified
[sinks.file]
enabled = true
path = "deploy/logs/app.log"
batch_size = 100
batch_timeout_ms = 1000 # Default is 8192 bytes if not specified
[[sinks]]
type = "File"
path = "deploy/logs/rustfs.log"
buffer_size = 102 # Default is 8192 bytes if not specified
flush_interval_ms = 1000
flush_threshold = 100
[logger]
queue_capacity = 10000
+103 -47
View File
@@ -1,6 +1,6 @@
use crate::global::{ENVIRONMENT, LOGGER_LEVEL, METER_INTERVAL, SAMPLE_RATIO, SERVICE_NAME, SERVICE_VERSION, USE_STDOUT};
use config::{Config, Environment, File, FileFormat};
use serde::Deserialize;
use config::{Config, File, FileFormat};
use rustfs_config::{APP_NAME, DEFAULT_LOG_LEVEL, ENVIRONMENT, METER_INTERVAL, SAMPLE_RATIO, SERVICE_VERSION, USE_STDOUT};
use serde::{Deserialize, Serialize};
use std::env;
/// OpenTelemetry Configuration
@@ -11,7 +11,7 @@ use std::env;
/// Add use_stdout for output to stdout
/// Add logger level for log level
/// Add local_logging_enabled for local logging enabled
#[derive(Debug, Deserialize, Clone)]
#[derive(Debug, Deserialize, Serialize, Clone)]
pub struct OtelConfig {
pub endpoint: String, // Endpoint for metric collection
pub use_stdout: Option<bool>, // Output to stdout
@@ -24,7 +24,7 @@ pub struct OtelConfig {
pub local_logging_enabled: Option<bool>, // Local logging enabled
}
// Helper function: Extract observable configuration from environment variables
/// Helper function: Extract observable configuration from environment variables
fn extract_otel_config_from_env() -> OtelConfig {
OtelConfig {
endpoint: env::var("RUSTFS_OBSERVABILITY_ENDPOINT").unwrap_or_else(|_| "".to_string()),
@@ -43,7 +43,7 @@ fn extract_otel_config_from_env() -> OtelConfig {
service_name: env::var("RUSTFS_OBSERVABILITY_SERVICE_NAME")
.ok()
.and_then(|v| v.parse().ok())
.or(Some(SERVICE_NAME.to_string())),
.or(Some(APP_NAME.to_string())),
service_version: env::var("RUSTFS_OBSERVABILITY_SERVICE_VERSION")
.ok()
.and_then(|v| v.parse().ok())
@@ -55,7 +55,7 @@ fn extract_otel_config_from_env() -> OtelConfig {
logger_level: env::var("RUSTFS_OBSERVABILITY_LOGGER_LEVEL")
.ok()
.and_then(|v| v.parse().ok())
.or(Some(LOGGER_LEVEL.to_string())),
.or(Some(DEFAULT_LOG_LEVEL.to_string())),
local_logging_enabled: env::var("RUSTFS_OBSERVABILITY_LOCAL_LOGGING_ENABLED")
.ok()
.and_then(|v| v.parse().ok())
@@ -63,36 +63,89 @@ fn extract_otel_config_from_env() -> OtelConfig {
}
}
impl Default for OtelConfig {
fn default() -> Self {
impl OtelConfig {
/// Create a new instance of OtelConfig with default values
///
/// # Returns
/// A new instance of OtelConfig
pub fn new() -> Self {
extract_otel_config_from_env()
}
}
impl Default for OtelConfig {
fn default() -> Self {
Self::new()
}
}
/// Kafka Sink Configuration - Add batch parameters
#[derive(Debug, Deserialize, Clone, Default)]
#[derive(Debug, Deserialize, Serialize, Clone)]
pub struct KafkaSinkConfig {
pub enabled: bool,
pub bootstrap_servers: String,
pub brokers: String,
pub topic: String,
pub batch_size: Option<usize>, // Batch size, default 100
pub batch_timeout_ms: Option<u64>, // Batch timeout time, default 1000ms
}
impl KafkaSinkConfig {
pub fn new() -> Self {
Self {
brokers: env::var("RUSTFS_SINKS_KAFKA_BROKERS")
.ok()
.filter(|s| !s.trim().is_empty())
.unwrap_or_else(|| "localhost:9092".to_string()),
topic: env::var("RUSTFS_SINKS_KAFKA_TOPIC")
.ok()
.filter(|s| !s.trim().is_empty())
.unwrap_or_else(|| "default_topic".to_string()),
batch_size: Some(100),
batch_timeout_ms: Some(1000),
}
}
}
impl Default for KafkaSinkConfig {
fn default() -> Self {
Self::new()
}
}
/// Webhook Sink Configuration - Add Retry Parameters
#[derive(Debug, Deserialize, Clone, Default)]
#[derive(Debug, Deserialize, Serialize, Clone)]
pub struct WebhookSinkConfig {
pub enabled: bool,
pub endpoint: String,
pub auth_token: String,
pub max_retries: Option<usize>, // Maximum number of retry times, default 3
pub retry_delay_ms: Option<u64>, // Retry the delay cardinality, default 100ms
}
impl WebhookSinkConfig {
pub fn new() -> Self {
Self {
endpoint: env::var("RUSTFS_SINKS_WEBHOOK_ENDPOINT")
.ok()
.filter(|s| !s.trim().is_empty())
.unwrap_or_else(|| "http://localhost:8080".to_string()),
auth_token: env::var("RUSTFS_SINKS_WEBHOOK_AUTH_TOKEN")
.ok()
.filter(|s| !s.trim().is_empty())
.unwrap_or_else(|| "default_token".to_string()),
max_retries: Some(3),
retry_delay_ms: Some(100),
}
}
}
impl Default for WebhookSinkConfig {
fn default() -> Self {
Self::new()
}
}
/// File Sink Configuration - Add buffering parameters
#[derive(Debug, Deserialize, Clone)]
#[derive(Debug, Deserialize, Serialize, Clone)]
pub struct FileSinkConfig {
pub enabled: bool,
pub path: String,
pub buffer_size: Option<usize>, // Write buffer size, default 8192
pub flush_interval_ms: Option<u64>, // Refresh interval time, default 1000ms
@@ -107,19 +160,15 @@ impl FileSinkConfig {
eprintln!("Failed to create log directory: {}", e);
return "rustfs/rustfs.log".to_string();
}
println!("Using log directory: {:?}", temp_dir);
temp_dir
.join("rustfs.log")
.to_str()
.unwrap_or("rustfs/rustfs.log")
.to_string()
}
}
impl Default for FileSinkConfig {
fn default() -> Self {
FileSinkConfig {
enabled: true,
pub fn new() -> Self {
Self {
path: env::var("RUSTFS_SINKS_FILE_PATH")
.ok()
.filter(|s| !s.trim().is_empty())
@@ -131,38 +180,53 @@ impl Default for FileSinkConfig {
}
}
impl Default for FileSinkConfig {
fn default() -> Self {
Self::new()
}
}
/// Sink configuration collection
#[derive(Debug, Deserialize, Clone)]
pub struct SinkConfig {
pub kafka: Option<KafkaSinkConfig>,
pub webhook: Option<WebhookSinkConfig>,
pub file: Option<FileSinkConfig>,
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(tag = "type")]
pub enum SinkConfig {
File(FileSinkConfig),
Kafka(KafkaSinkConfig),
Webhook(WebhookSinkConfig),
}
impl SinkConfig {
pub fn new() -> Self {
Self::File(FileSinkConfig::new())
}
}
impl Default for SinkConfig {
fn default() -> Self {
SinkConfig {
kafka: None,
webhook: None,
file: Some(FileSinkConfig::default()),
}
Self::new()
}
}
///Logger Configuration
#[derive(Debug, Deserialize, Clone)]
#[derive(Debug, Deserialize, Serialize, Clone)]
pub struct LoggerConfig {
pub queue_capacity: Option<usize>,
}
impl Default for LoggerConfig {
fn default() -> Self {
LoggerConfig {
impl LoggerConfig {
pub fn new() -> Self {
Self {
queue_capacity: Some(10000),
}
}
}
impl Default for LoggerConfig {
fn default() -> Self {
Self::new()
}
}
/// Overall application configuration
/// Add observability, sinks, and logger configuration
///
@@ -180,7 +244,7 @@ impl Default for LoggerConfig {
#[derive(Debug, Deserialize, Clone)]
pub struct AppConfig {
pub observability: OtelConfig,
pub sinks: SinkConfig,
pub sinks: Vec<SinkConfig>,
pub logger: Option<LoggerConfig>,
}
@@ -192,7 +256,7 @@ impl AppConfig {
pub fn new() -> Self {
Self {
observability: OtelConfig::default(),
sinks: SinkConfig::default(),
sinks: vec![SinkConfig::default()],
logger: Some(LoggerConfig::default()),
}
}
@@ -258,14 +322,6 @@ pub fn load_config(config_dir: Option<String>) -> AppConfig {
let app_config = Config::builder()
.add_source(File::with_name(config_dir.as_str()).format(FileFormat::Toml).required(false))
.add_source(File::with_name(config_dir.as_str()).format(FileFormat::Yaml).required(false))
.add_source(
Environment::default()
.prefix("RUSTFS")
.prefix_separator("__")
.separator("__")
.with_list_parse_key("volumes")
.try_parsing(true),
)
.build()
.unwrap_or_default();
-8
View File
@@ -3,14 +3,6 @@ use std::sync::{Arc, Mutex};
use tokio::sync::{OnceCell, SetError};
use tracing::{error, info};
pub(crate) const USE_STDOUT: bool = true;
pub(crate) const SERVICE_NAME: &str = "RustFS";
pub(crate) const SAMPLE_RATIO: f64 = 1.0;
pub(crate) const METER_INTERVAL: u64 = 60;
pub(crate) const SERVICE_VERSION: &str = "0.1.0";
pub(crate) const ENVIRONMENT: &str = "production";
pub(crate) const LOGGER_LEVEL: &str = "info";
/// Global guard for OpenTelemetry tracing
static GLOBAL_GUARD: OnceCell<Arc<Mutex<OtelGuard>>> = OnceCell::const_new();
+2 -9
View File
@@ -32,20 +32,13 @@ mod config;
mod entry;
mod global;
mod logger;
mod sink;
mod sinks;
mod system;
mod telemetry;
mod utils;
mod worker;
use crate::logger::InitLogStatus;
pub use config::load_config;
#[cfg(feature = "file")]
pub use config::FileSinkConfig;
#[cfg(feature = "kafka")]
pub use config::KafkaSinkConfig;
#[cfg(feature = "webhook")]
pub use config::WebhookSinkConfig;
pub use config::{AppConfig, LoggerConfig, OtelConfig, SinkConfig};
pub use entry::args::Args;
pub use entry::audit::{ApiDetails, AuditLogEntry};
@@ -79,7 +72,7 @@ use tracing::{error, info};
/// ```
pub async fn init_obs(config: AppConfig) -> (Arc<Mutex<Logger>>, telemetry::OtelGuard) {
let guard = init_telemetry(&config.observability);
let sinks = sink::create_sinks(&config).await;
let sinks = sinks::create_sinks(&config).await;
let logger = init_global_logger(&config, sinks).await;
let obs_config = config.observability.clone();
tokio::spawn(async move {
+4 -4
View File
@@ -1,6 +1,6 @@
use crate::global::{ENVIRONMENT, SERVICE_NAME, SERVICE_VERSION};
use crate::sink::Sink;
use crate::sinks::Sink;
use crate::{AppConfig, AuditLogEntry, BaseLogEntry, ConsoleLogEntry, GlobalError, OtelConfig, ServerLogEntry, UnifiedLogEntry};
use rustfs_config::{APP_NAME, ENVIRONMENT, SERVICE_VERSION};
use std::sync::Arc;
use std::time::SystemTime;
use tokio::sync::mpsc::{self, Receiver, Sender};
@@ -428,7 +428,7 @@ impl Default for InitLogStatus {
fn default() -> Self {
Self {
timestamp: SystemTime::now(),
service_name: String::from(SERVICE_NAME),
service_name: String::from(APP_NAME),
version: SERVICE_VERSION.to_string(),
environment: ENVIRONMENT.to_string(),
}
@@ -442,7 +442,7 @@ impl InitLogStatus {
let version = config.service_version.unwrap_or(SERVICE_VERSION.to_string());
Self {
timestamp: SystemTime::now(),
service_name: String::from(SERVICE_NAME),
service_name: String::from(APP_NAME),
version,
environment,
}
-497
View File
@@ -1,497 +0,0 @@
use crate::{AppConfig, LogRecord, UnifiedLogEntry};
use async_trait::async_trait;
use std::sync::Arc;
use tokio::fs::OpenOptions;
use tokio::io;
use tokio::io::AsyncWriteExt;
/// Sink Trait definition, asynchronously write logs
#[async_trait]
pub trait Sink: Send + Sync {
async fn write(&self, entry: &UnifiedLogEntry);
}
#[cfg(feature = "kafka")]
/// Kafka Sink Implementation
pub struct KafkaSink {
producer: rdkafka::producer::FutureProducer,
topic: String,
batch_size: usize,
batch_timeout_ms: u64,
entries: Arc<tokio::sync::Mutex<Vec<UnifiedLogEntry>>>,
last_flush: Arc<std::sync::atomic::AtomicU64>,
}
#[cfg(feature = "kafka")]
impl KafkaSink {
/// Create a new KafkaSink instance
pub fn new(producer: rdkafka::producer::FutureProducer, topic: String, batch_size: usize, batch_timeout_ms: u64) -> Self {
// Create Arc-wrapped values first
let entries = Arc::new(tokio::sync::Mutex::new(Vec::with_capacity(batch_size)));
let last_flush = Arc::new(std::sync::atomic::AtomicU64::new(
std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64,
));
let sink = KafkaSink {
producer: producer.clone(),
topic: topic.clone(),
batch_size,
batch_timeout_ms,
entries: entries.clone(),
last_flush: last_flush.clone(),
};
// Start background flusher
tokio::spawn(Self::periodic_flush(producer, topic, entries, last_flush, batch_timeout_ms));
sink
}
/// Add a getter method to read the batch_timeout_ms field
#[allow(dead_code)]
pub fn batch_timeout(&self) -> u64 {
self.batch_timeout_ms
}
/// Add a method to dynamically adjust the timeout if needed
#[allow(dead_code)]
pub fn set_batch_timeout(&mut self, new_timeout_ms: u64) {
self.batch_timeout_ms = new_timeout_ms;
}
async fn periodic_flush(
producer: rdkafka::producer::FutureProducer,
topic: String,
entries: Arc<tokio::sync::Mutex<Vec<UnifiedLogEntry>>>,
last_flush: Arc<std::sync::atomic::AtomicU64>,
timeout_ms: u64,
) {
loop {
tokio::time::sleep(tokio::time::Duration::from_millis(timeout_ms / 2)).await;
let now = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64;
let last = last_flush.load(std::sync::atomic::Ordering::Relaxed);
if now - last >= timeout_ms {
let mut batch = entries.lock().await;
if !batch.is_empty() {
Self::send_batch(&producer, &topic, batch.drain(..).collect()).await;
last_flush.store(now, std::sync::atomic::Ordering::Relaxed);
}
}
}
}
async fn send_batch(producer: &rdkafka::producer::FutureProducer, topic: &str, entries: Vec<UnifiedLogEntry>) {
for entry in entries {
let payload = match serde_json::to_string(&entry) {
Ok(p) => p,
Err(e) => {
eprintln!("Failed to serialize log entry: {}", e);
continue;
}
};
let span_id = entry.get_timestamp().to_rfc3339();
let _ = producer
.send(
rdkafka::producer::FutureRecord::to(topic).payload(&payload).key(&span_id),
std::time::Duration::from_secs(5),
)
.await;
}
}
}
#[cfg(feature = "kafka")]
#[async_trait]
impl Sink for KafkaSink {
async fn write(&self, entry: &UnifiedLogEntry) {
let mut batch = self.entries.lock().await;
batch.push(entry.clone());
let should_flush_by_size = batch.len() >= self.batch_size;
let should_flush_by_time = {
let now = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64;
let last = self.last_flush.load(std::sync::atomic::Ordering::Relaxed);
now - last >= self.batch_timeout_ms
};
if should_flush_by_size || should_flush_by_time {
// Existing flush logic
let entries_to_send: Vec<UnifiedLogEntry> = batch.drain(..).collect();
let producer = self.producer.clone();
let topic = self.topic.clone();
self.last_flush.store(
std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64,
std::sync::atomic::Ordering::Relaxed,
);
tokio::spawn(async move {
KafkaSink::send_batch(&producer, &topic, entries_to_send).await;
});
}
}
}
#[cfg(feature = "kafka")]
impl Drop for KafkaSink {
fn drop(&mut self) {
// Perform any necessary cleanup here
// For example, you might want to flush any remaining entries
let producer = self.producer.clone();
let topic = self.topic.clone();
let entries = self.entries.clone();
let last_flush = self.last_flush.clone();
tokio::spawn(async move {
let mut batch = entries.lock().await;
if !batch.is_empty() {
KafkaSink::send_batch(&producer, &topic, batch.drain(..).collect()).await;
last_flush.store(
std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64,
std::sync::atomic::Ordering::Relaxed,
);
}
});
eprintln!("Dropping KafkaSink with topic: {}", self.topic);
}
}
#[cfg(feature = "webhook")]
/// Webhook Sink Implementation
pub struct WebhookSink {
endpoint: String,
auth_token: String,
client: reqwest::Client,
max_retries: usize,
retry_delay_ms: u64,
}
#[cfg(feature = "webhook")]
impl WebhookSink {
pub fn new(endpoint: String, auth_token: String, max_retries: usize, retry_delay_ms: u64) -> Self {
WebhookSink {
endpoint,
auth_token,
client: reqwest::Client::builder()
.timeout(std::time::Duration::from_secs(10))
.build()
.unwrap_or_else(|_| reqwest::Client::new()),
max_retries,
retry_delay_ms,
}
}
}
#[cfg(feature = "webhook")]
#[async_trait]
impl Sink for WebhookSink {
async fn write(&self, entry: &UnifiedLogEntry) {
let mut retries = 0;
let url = self.endpoint.clone();
let entry_clone = entry.clone();
let auth_value = reqwest::header::HeaderValue::from_str(format!("Bearer {}", self.auth_token.clone()).as_str()).unwrap();
while retries < self.max_retries {
match self
.client
.post(&url)
.header(reqwest::header::AUTHORIZATION, auth_value.clone())
.json(&entry_clone)
.send()
.await
{
Ok(response) if response.status().is_success() => {
return;
}
_ => {
retries += 1;
if retries < self.max_retries {
tokio::time::sleep(tokio::time::Duration::from_millis(
self.retry_delay_ms * (1 << retries), // Exponential backoff
))
.await;
}
}
}
}
eprintln!("Failed to send log to webhook after {} retries", self.max_retries);
}
}
#[cfg(feature = "webhook")]
impl Drop for WebhookSink {
fn drop(&mut self) {
// Perform any necessary cleanup here
// For example, you might want to log that the sink is being dropped
eprintln!("Dropping WebhookSink with URL: {}", self.endpoint);
}
}
#[cfg(feature = "file")]
/// File Sink Implementation
pub struct FileSink {
path: String,
buffer_size: usize,
writer: Arc<tokio::sync::Mutex<io::BufWriter<tokio::fs::File>>>,
entry_count: std::sync::atomic::AtomicUsize,
last_flush: std::sync::atomic::AtomicU64,
flush_interval_ms: u64, // Time between flushes
flush_threshold: usize, // Number of entries before flush
}
#[cfg(feature = "file")]
impl FileSink {
/// Create a new FileSink instance
pub async fn new(
path: String,
buffer_size: usize,
flush_interval_ms: u64,
flush_threshold: usize,
) -> Result<Self, io::Error> {
// check if the file exists
let file_exists = tokio::fs::metadata(&path).await.is_ok();
// if the file not exists, create it
if !file_exists {
tokio::fs::create_dir_all(std::path::Path::new(&path).parent().unwrap()).await?;
tracing::debug!("File does not exist, creating it. Path: {:?}", path)
}
let file = if file_exists {
// If the file exists, open it in append mode
tracing::debug!("FileSink: File exists, opening in append mode.");
OpenOptions::new().append(true).create(true).open(&path).await?
} else {
// If the file does not exist, create it
tracing::debug!("FileSink: File does not exist, creating a new file.");
// Create the file and write a header or initial content if needed
OpenOptions::new().create(true).truncate(true).write(true).open(&path).await?
};
let writer = io::BufWriter::with_capacity(buffer_size, file);
let now = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64;
Ok(FileSink {
path,
buffer_size,
writer: Arc::new(tokio::sync::Mutex::new(writer)),
entry_count: std::sync::atomic::AtomicUsize::new(0),
last_flush: std::sync::atomic::AtomicU64::new(now),
flush_interval_ms,
flush_threshold,
})
}
#[allow(dead_code)]
async fn initialize_writer(&mut self) -> io::Result<()> {
let file = tokio::fs::File::create(&self.path).await?;
// Use buffer_size to create a buffer writer with a specified capacity
let buf_writer = io::BufWriter::with_capacity(self.buffer_size, file);
// Replace the original writer with the new Mutex
self.writer = Arc::new(tokio::sync::Mutex::new(buf_writer));
Ok(())
}
// Get the current buffer size
#[allow(dead_code)]
pub fn buffer_size(&self) -> usize {
self.buffer_size
}
// How to dynamically adjust the buffer size
#[allow(dead_code)]
pub async fn set_buffer_size(&mut self, new_size: usize) -> io::Result<()> {
if self.buffer_size != new_size {
self.buffer_size = new_size;
// Reinitialize the writer directly, without checking is_some()
self.initialize_writer().await?;
}
Ok(())
}
// Check if flushing is needed based on count or time
fn should_flush(&self) -> bool {
// Check entry count threshold
if self.entry_count.load(std::sync::atomic::Ordering::Relaxed) >= self.flush_threshold {
return true;
}
// Check time threshold
let now = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64;
let last = self.last_flush.load(std::sync::atomic::Ordering::Relaxed);
now - last >= self.flush_interval_ms
}
}
#[cfg(feature = "file")]
#[async_trait]
impl Sink for FileSink {
async fn write(&self, entry: &UnifiedLogEntry) {
let line = format!("{:?}\n", entry);
let mut writer = self.writer.lock().await;
if let Err(e) = writer.write_all(line.as_bytes()).await {
eprintln!(
"Failed to write log to file {}: {},entry timestamp:{:?}",
self.path,
e,
entry.get_timestamp()
);
return;
}
// Only flush periodically to improve performance
// Logic to determine when to flush could be added here
// Increment the entry count
self.entry_count.fetch_add(1, std::sync::atomic::Ordering::Relaxed);
// Check if we should flush
if self.should_flush() {
if let Err(e) = writer.flush().await {
eprintln!("Failed to flush log file {}: {}", self.path, e);
return;
}
// Reset counters
self.entry_count.store(0, std::sync::atomic::Ordering::Relaxed);
let now = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64;
self.last_flush.store(now, std::sync::atomic::Ordering::Relaxed);
}
}
}
#[cfg(feature = "file")]
impl Drop for FileSink {
fn drop(&mut self) {
let writer = self.writer.clone();
let path = self.path.clone();
tokio::task::spawn_blocking(move || {
let rt = tokio::runtime::Runtime::new().unwrap();
rt.block_on(async {
let mut writer = writer.lock().await;
if let Err(e) = writer.flush().await {
eprintln!("Failed to flush log file {}: {}", path, e);
}
});
});
}
}
/// Create a list of Sink instances
pub async fn create_sinks(config: &AppConfig) -> Vec<Arc<dyn Sink>> {
let mut sinks: Vec<Arc<dyn Sink>> = Vec::new();
#[cfg(feature = "kafka")]
{
match &config.sinks.kafka {
Some(sink_kafka) => {
if sink_kafka.enabled {
match rdkafka::config::ClientConfig::new()
.set("bootstrap.servers", &sink_kafka.bootstrap_servers)
.set("message.timeout.ms", "5000")
.create()
{
Ok(producer) => {
sinks.push(Arc::new(KafkaSink::new(
producer,
sink_kafka.topic.clone(),
sink_kafka.batch_size.unwrap_or(100),
sink_kafka.batch_timeout_ms.unwrap_or(1000),
)));
}
Err(e) => {
tracing::error!("Failed to create Kafka producer: {}", e);
}
}
} else {
tracing::info!("Kafka sink is disabled in the configuration");
}
}
_ => {
tracing::info!("Kafka sink is not configured or disabled");
}
}
}
#[cfg(feature = "webhook")]
{
match &config.sinks.webhook {
Some(sink_webhook) => {
if sink_webhook.enabled {
sinks.push(Arc::new(WebhookSink::new(
sink_webhook.endpoint.clone(),
sink_webhook.auth_token.clone(),
sink_webhook.max_retries.unwrap_or(3),
sink_webhook.retry_delay_ms.unwrap_or(100),
)));
} else {
tracing::info!("Webhook sink is disabled in the configuration");
}
}
_ => {
tracing::info!("Webhook sink is not configured or disabled");
}
}
}
#[cfg(feature = "file")]
{
// let config = config.clone();
match &config.sinks.file {
Some(sink_file) => {
tracing::info!("File sink is enabled in the configuration");
let path = if sink_file.enabled {
sink_file.path.clone()
} else {
"rustfs.log".to_string()
};
tracing::debug!("FileSink: Using path: {}", path);
sinks.push(Arc::new(
FileSink::new(
path.clone(),
sink_file.buffer_size.unwrap_or(8192),
sink_file.flush_interval_ms.unwrap_or(1000),
sink_file.flush_threshold.unwrap_or(100),
)
.await
.unwrap(),
));
}
_ => {
tracing::info!("File sink is not configured or disabled");
}
}
}
sinks
}
+164
View File
@@ -0,0 +1,164 @@
use crate::sinks::Sink;
use crate::{LogRecord, UnifiedLogEntry};
use async_trait::async_trait;
use std::sync::Arc;
use tokio::fs::OpenOptions;
use tokio::io;
use tokio::io::AsyncWriteExt;
/// File Sink Implementation
pub struct FileSink {
path: String,
buffer_size: usize,
writer: Arc<tokio::sync::Mutex<io::BufWriter<tokio::fs::File>>>,
entry_count: std::sync::atomic::AtomicUsize,
last_flush: std::sync::atomic::AtomicU64,
flush_interval_ms: u64, // Time between flushes
flush_threshold: usize, // Number of entries before flush
}
impl FileSink {
/// Create a new FileSink instance
pub async fn new(
path: String,
buffer_size: usize,
flush_interval_ms: u64,
flush_threshold: usize,
) -> Result<Self, io::Error> {
// check if the file exists
let file_exists = tokio::fs::metadata(&path).await.is_ok();
// if the file not exists, create it
if !file_exists {
tokio::fs::create_dir_all(std::path::Path::new(&path).parent().unwrap()).await?;
tracing::debug!("File does not exist, creating it. Path: {:?}", path)
}
let file = if file_exists {
// If the file exists, open it in append mode
tracing::debug!("FileSink: File exists, opening in append mode.");
OpenOptions::new().append(true).create(true).open(&path).await?
} else {
// If the file does not exist, create it
tracing::debug!("FileSink: File does not exist, creating a new file.");
// Create the file and write a header or initial content if needed
OpenOptions::new().create(true).truncate(true).write(true).open(&path).await?
};
let writer = io::BufWriter::with_capacity(buffer_size, file);
let now = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64;
Ok(FileSink {
path,
buffer_size,
writer: Arc::new(tokio::sync::Mutex::new(writer)),
entry_count: std::sync::atomic::AtomicUsize::new(0),
last_flush: std::sync::atomic::AtomicU64::new(now),
flush_interval_ms,
flush_threshold,
})
}
#[allow(dead_code)]
async fn initialize_writer(&mut self) -> io::Result<()> {
let file = tokio::fs::File::create(&self.path).await?;
// Use buffer_size to create a buffer writer with a specified capacity
let buf_writer = io::BufWriter::with_capacity(self.buffer_size, file);
// Replace the original writer with the new Mutex
self.writer = Arc::new(tokio::sync::Mutex::new(buf_writer));
Ok(())
}
// Get the current buffer size
#[allow(dead_code)]
pub fn buffer_size(&self) -> usize {
self.buffer_size
}
// How to dynamically adjust the buffer size
#[allow(dead_code)]
pub async fn set_buffer_size(&mut self, new_size: usize) -> io::Result<()> {
if self.buffer_size != new_size {
self.buffer_size = new_size;
// Reinitialize the writer directly, without checking is_some()
self.initialize_writer().await?;
}
Ok(())
}
// Check if flushing is needed based on count or time
fn should_flush(&self) -> bool {
// Check entry count threshold
if self.entry_count.load(std::sync::atomic::Ordering::Relaxed) >= self.flush_threshold {
return true;
}
// Check time threshold
let now = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64;
let last = self.last_flush.load(std::sync::atomic::Ordering::Relaxed);
now - last >= self.flush_interval_ms
}
}
#[async_trait]
impl Sink for FileSink {
async fn write(&self, entry: &UnifiedLogEntry) {
let line = format!("{:?}\n", entry);
let mut writer = self.writer.lock().await;
if let Err(e) = writer.write_all(line.as_bytes()).await {
eprintln!(
"Failed to write log to file {}: {},entry timestamp:{:?}",
self.path,
e,
entry.get_timestamp()
);
return;
}
// Only flush periodically to improve performance
// Logic to determine when to flush could be added here
// Increment the entry count
self.entry_count.fetch_add(1, std::sync::atomic::Ordering::Relaxed);
// Check if we should flush
if self.should_flush() {
if let Err(e) = writer.flush().await {
eprintln!("Failed to flush log file {}: {}", self.path, e);
return;
}
// Reset counters
self.entry_count.store(0, std::sync::atomic::Ordering::Relaxed);
let now = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64;
self.last_flush.store(now, std::sync::atomic::Ordering::Relaxed);
}
}
}
impl Drop for FileSink {
fn drop(&mut self) {
let writer = self.writer.clone();
let path = self.path.clone();
tokio::task::spawn_blocking(move || {
let rt = tokio::runtime::Runtime::new().unwrap();
rt.block_on(async {
let mut writer = writer.lock().await;
if let Err(e) = writer.flush().await {
eprintln!("Failed to flush log file {}: {}", path, e);
}
});
});
}
}
+165
View File
@@ -0,0 +1,165 @@
use crate::sinks::Sink;
use crate::{LogRecord, UnifiedLogEntry};
use async_trait::async_trait;
use std::sync::Arc;
/// Kafka Sink Implementation
pub struct KafkaSink {
producer: rdkafka::producer::FutureProducer,
topic: String,
batch_size: usize,
batch_timeout_ms: u64,
entries: Arc<tokio::sync::Mutex<Vec<UnifiedLogEntry>>>,
last_flush: Arc<std::sync::atomic::AtomicU64>,
}
impl KafkaSink {
/// Create a new KafkaSink instance
pub fn new(producer: rdkafka::producer::FutureProducer, topic: String, batch_size: usize, batch_timeout_ms: u64) -> Self {
// Create Arc-wrapped values first
let entries = Arc::new(tokio::sync::Mutex::new(Vec::with_capacity(batch_size)));
let last_flush = Arc::new(std::sync::atomic::AtomicU64::new(
std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64,
));
let sink = KafkaSink {
producer: producer.clone(),
topic: topic.clone(),
batch_size,
batch_timeout_ms,
entries: entries.clone(),
last_flush: last_flush.clone(),
};
// Start background flusher
tokio::spawn(Self::periodic_flush(producer, topic, entries, last_flush, batch_timeout_ms));
sink
}
/// Add a getter method to read the batch_timeout_ms field
#[allow(dead_code)]
pub fn batch_timeout(&self) -> u64 {
self.batch_timeout_ms
}
/// Add a method to dynamically adjust the timeout if needed
#[allow(dead_code)]
pub fn set_batch_timeout(&mut self, new_timeout_ms: u64) {
self.batch_timeout_ms = new_timeout_ms;
}
async fn periodic_flush(
producer: rdkafka::producer::FutureProducer,
topic: String,
entries: Arc<tokio::sync::Mutex<Vec<UnifiedLogEntry>>>,
last_flush: Arc<std::sync::atomic::AtomicU64>,
timeout_ms: u64,
) {
loop {
tokio::time::sleep(tokio::time::Duration::from_millis(timeout_ms / 2)).await;
let now = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64;
let last = last_flush.load(std::sync::atomic::Ordering::Relaxed);
if now - last >= timeout_ms {
let mut batch = entries.lock().await;
if !batch.is_empty() {
Self::send_batch(&producer, &topic, batch.drain(..).collect()).await;
last_flush.store(now, std::sync::atomic::Ordering::Relaxed);
}
}
}
}
async fn send_batch(producer: &rdkafka::producer::FutureProducer, topic: &str, entries: Vec<UnifiedLogEntry>) {
for entry in entries {
let payload = match serde_json::to_string(&entry) {
Ok(p) => p,
Err(e) => {
eprintln!("Failed to serialize log entry: {}", e);
continue;
}
};
let span_id = entry.get_timestamp().to_rfc3339();
let _ = producer
.send(
rdkafka::producer::FutureRecord::to(topic).payload(&payload).key(&span_id),
std::time::Duration::from_secs(5),
)
.await;
}
}
}
#[async_trait]
impl Sink for KafkaSink {
async fn write(&self, entry: &UnifiedLogEntry) {
let mut batch = self.entries.lock().await;
batch.push(entry.clone());
let should_flush_by_size = batch.len() >= self.batch_size;
let should_flush_by_time = {
let now = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64;
let last = self.last_flush.load(std::sync::atomic::Ordering::Relaxed);
now - last >= self.batch_timeout_ms
};
if should_flush_by_size || should_flush_by_time {
// Existing flush logic
let entries_to_send: Vec<UnifiedLogEntry> = batch.drain(..).collect();
let producer = self.producer.clone();
let topic = self.topic.clone();
self.last_flush.store(
std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64,
std::sync::atomic::Ordering::Relaxed,
);
tokio::spawn(async move {
KafkaSink::send_batch(&producer, &topic, entries_to_send).await;
});
}
}
}
impl Drop for KafkaSink {
fn drop(&mut self) {
// Perform any necessary cleanup here
// For example, you might want to flush any remaining entries
let producer = self.producer.clone();
let topic = self.topic.clone();
let entries = self.entries.clone();
let last_flush = self.last_flush.clone();
tokio::spawn(async move {
let mut batch = entries.lock().await;
if !batch.is_empty() {
KafkaSink::send_batch(&producer, &topic, batch.drain(..).collect()).await;
last_flush.store(
std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.unwrap()
.as_millis() as u64,
std::sync::atomic::Ordering::Relaxed,
);
}
});
eprintln!("Dropping KafkaSink with topic: {}", self.topic);
}
}
+92
View File
@@ -0,0 +1,92 @@
use crate::{AppConfig, SinkConfig, UnifiedLogEntry};
use async_trait::async_trait;
use std::sync::Arc;
#[cfg(feature = "file")]
mod file;
#[cfg(all(feature = "kafka", target_os = "linux"))]
mod kafka;
#[cfg(feature = "webhook")]
mod webhook;
/// Sink Trait definition, asynchronously write logs
#[async_trait]
pub trait Sink: Send + Sync {
async fn write(&self, entry: &UnifiedLogEntry);
}
/// Create a list of Sink instances
pub async fn create_sinks(config: &AppConfig) -> Vec<Arc<dyn Sink>> {
let mut sinks: Vec<Arc<dyn Sink>> = Vec::new();
for sink_config in &config.sinks {
match sink_config {
#[cfg(all(feature = "kafka", target_os = "linux"))]
SinkConfig::Kafka(kafka_config) => {
match rdkafka::config::ClientConfig::new()
.set("bootstrap.servers", &kafka_config.brokers)
.set("message.timeout.ms", "5000")
.create()
{
Ok(producer) => {
sinks.push(Arc::new(kafka::KafkaSink::new(
producer,
kafka_config.topic.clone(),
kafka_config.batch_size.unwrap_or(100),
kafka_config.batch_timeout_ms.unwrap_or(1000),
)));
tracing::info!("Kafka sink created for topic: {}", kafka_config.topic);
}
Err(e) => {
tracing::error!("Failed to create Kafka producer: {}", e);
}
}
}
#[cfg(feature = "webhook")]
SinkConfig::Webhook(webhook_config) => {
sinks.push(Arc::new(webhook::WebhookSink::new(
webhook_config.endpoint.clone(),
webhook_config.auth_token.clone(),
webhook_config.max_retries.unwrap_or(3),
webhook_config.retry_delay_ms.unwrap_or(100),
)));
tracing::info!("Webhook sink created for endpoint: {}", webhook_config.endpoint);
}
#[cfg(feature = "file")]
SinkConfig::File(file_config) => {
tracing::debug!("FileSink: Using path: {}", file_config.path);
match file::FileSink::new(
file_config.path.clone(),
file_config.buffer_size.unwrap_or(8192),
file_config.flush_interval_ms.unwrap_or(1000),
file_config.flush_threshold.unwrap_or(100),
)
.await
{
Ok(sink) => {
sinks.push(Arc::new(sink));
tracing::info!("File sink created for path: {}", file_config.path);
}
Err(e) => {
tracing::error!("Failed to create File sink: {}", e);
}
}
}
#[cfg(any(not(feature = "kafka"), not(target_os = "linux")))]
SinkConfig::Kafka(_) => {
tracing::warn!("Kafka sink is configured but the 'kafka' feature is not enabled");
}
#[cfg(not(feature = "webhook"))]
SinkConfig::Webhook(_) => {
tracing::warn!("Webhook sink is configured but the 'webhook' feature is not enabled");
}
#[cfg(not(feature = "file"))]
SinkConfig::File(_) => {
tracing::warn!("File sink is configured but the 'file' feature is not enabled");
}
}
}
sinks
}
+70
View File
@@ -0,0 +1,70 @@
use crate::sinks::Sink;
use crate::UnifiedLogEntry;
use async_trait::async_trait;
/// Webhook Sink Implementation
pub struct WebhookSink {
endpoint: String,
auth_token: String,
client: reqwest::Client,
max_retries: usize,
retry_delay_ms: u64,
}
impl WebhookSink {
pub fn new(endpoint: String, auth_token: String, max_retries: usize, retry_delay_ms: u64) -> Self {
WebhookSink {
endpoint,
auth_token,
client: reqwest::Client::builder()
.timeout(std::time::Duration::from_secs(10))
.build()
.unwrap_or_else(|_| reqwest::Client::new()),
max_retries,
retry_delay_ms,
}
}
}
#[async_trait]
impl Sink for WebhookSink {
async fn write(&self, entry: &UnifiedLogEntry) {
let mut retries = 0;
let url = self.endpoint.clone();
let entry_clone = entry.clone();
let auth_value = reqwest::header::HeaderValue::from_str(format!("Bearer {}", self.auth_token.clone()).as_str()).unwrap();
while retries < self.max_retries {
match self
.client
.post(&url)
.header(reqwest::header::AUTHORIZATION, auth_value.clone())
.json(&entry_clone)
.send()
.await
{
Ok(response) if response.status().is_success() => {
return;
}
_ => {
retries += 1;
if retries < self.max_retries {
tokio::time::sleep(tokio::time::Duration::from_millis(
self.retry_delay_ms * (1 << retries), // Exponential backoff
))
.await;
}
}
}
}
eprintln!("Failed to send log to webhook after {} retries", self.max_retries);
}
}
impl Drop for WebhookSink {
fn drop(&mut self) {
// Perform any necessary cleanup here
// For example, you might want to log that the sink is being dropped
eprintln!("Dropping WebhookSink with URL: {}", self.endpoint);
}
}
+13 -10
View File
@@ -1,5 +1,3 @@
use crate::global::{ENVIRONMENT, LOGGER_LEVEL, METER_INTERVAL, SAMPLE_RATIO, SERVICE_NAME, SERVICE_VERSION, USE_STDOUT};
use crate::utils::get_local_ip_with_default;
use crate::OtelConfig;
use opentelemetry::trace::TracerProvider;
use opentelemetry::{global, KeyValue};
@@ -15,6 +13,8 @@ use opentelemetry_semantic_conventions::{
attribute::{DEPLOYMENT_ENVIRONMENT_NAME, NETWORK_LOCAL_ADDRESS, SERVICE_VERSION as OTEL_SERVICE_VERSION},
SCHEMA_URL,
};
use rustfs_config::{APP_NAME, DEFAULT_LOG_LEVEL, ENVIRONMENT, METER_INTERVAL, SAMPLE_RATIO, SERVICE_VERSION, USE_STDOUT};
use rustfs_utils::get_local_ip_with_default;
use smallvec::SmallVec;
use std::borrow::Cow;
use std::io::IsTerminal;
@@ -70,7 +70,7 @@ impl Drop for OtelGuard {
/// create OpenTelemetry Resource
fn resource(config: &OtelConfig) -> Resource {
Resource::builder()
.with_service_name(Cow::Borrowed(config.service_name.as_deref().unwrap_or(SERVICE_NAME)).to_string())
.with_service_name(Cow::Borrowed(config.service_name.as_deref().unwrap_or(APP_NAME)).to_string())
.with_schema_url(
[
KeyValue::new(
@@ -101,8 +101,8 @@ pub fn init_telemetry(config: &OtelConfig) -> OtelGuard {
let endpoint = &config.endpoint;
let use_stdout = config.use_stdout.unwrap_or(USE_STDOUT);
let meter_interval = config.meter_interval.unwrap_or(METER_INTERVAL);
let logger_level = config.logger_level.as_deref().unwrap_or(LOGGER_LEVEL);
let service_name = config.service_name.as_deref().unwrap_or(SERVICE_NAME);
let logger_level = config.logger_level.as_deref().unwrap_or(DEFAULT_LOG_LEVEL);
let service_name = config.service_name.as_deref().unwrap_or(APP_NAME);
// Pre-create resource objects to avoid repeated construction
let res = resource(config);
@@ -210,7 +210,8 @@ pub fn init_telemetry(config: &OtelConfig) -> OtelGuard {
.with_thread_names(true)
.with_thread_ids(true)
.with_file(true)
.with_line_number(true);
.with_line_number(true)
.with_filter(build_env_filter(logger_level, None));
let filter = build_env_filter(logger_level, None);
let otel_filter = build_env_filter(logger_level, None);
@@ -218,7 +219,7 @@ pub fn init_telemetry(config: &OtelConfig) -> OtelGuard {
let tracer = tracer_provider.tracer(Cow::Borrowed(service_name).to_string());
// Configure registry to avoid repeated calls to filter methods
let _registry = tracing_subscriber::registry()
tracing_subscriber::registry()
.with(filter)
.with(ErrorLayer::default())
.with(if config.local_logging_enabled.unwrap_or(false) {
@@ -230,10 +231,13 @@ pub fn init_telemetry(config: &OtelConfig) -> OtelGuard {
.with(otel_layer)
.with(MetricsLayer::new(meter_provider.clone()))
.init();
if !endpoint.is_empty() {
info!(
"OpenTelemetry telemetry initialized with OTLP endpoint: {}, logger_level: {}",
endpoint, logger_level
"OpenTelemetry telemetry initialized with OTLP endpoint: {}, logger_level: {},RUST_LOG env: {}",
endpoint,
logger_level,
std::env::var("RUST_LOG").unwrap_or_else(|_| "未设置".to_string())
);
}
}
@@ -248,7 +252,6 @@ pub fn init_telemetry(config: &OtelConfig) -> OtelGuard {
fn build_env_filter(logger_level: &str, default_level: Option<&str>) -> EnvFilter {
let level = default_level.unwrap_or(logger_level);
let mut filter = EnvFilter::try_from_default_env().unwrap_or_else(|_| EnvFilter::new(level));
if !matches!(logger_level, "trace" | "debug") {
let directives: SmallVec<[&str; 5]> = smallvec::smallvec!["hyper", "tonic", "h2", "reqwest", "tower"];
for directive in directives {
+2 -2
View File
@@ -1,9 +1,9 @@
use crate::{sink::Sink, UnifiedLogEntry};
use crate::{sinks::Sink, UnifiedLogEntry};
use std::sync::Arc;
use tokio::sync::mpsc::Receiver;
/// Start the log processing worker thread
pub async fn start_worker(receiver: Receiver<UnifiedLogEntry>, sinks: Vec<Arc<dyn Sink>>) {
pub(crate) async fn start_worker(receiver: Receiver<UnifiedLogEntry>, sinks: Vec<Arc<dyn Sink>>) {
let mut receiver = receiver;
while let Some(entry) = receiver.recv().await {
for sink in &sinks {
+25
View File
@@ -0,0 +1,25 @@
[package]
name = "rustfs-utils"
edition.workspace = true
license.workspace = true
repository.workspace = true
rust-version.workspace = true
version.workspace = true
[dependencies]
local-ip-address = { workspace = true, optional = true }
rustfs-config = { workspace = true }
rustls = { workspace = true, optional = true }
rustls-pemfile = { workspace = true, optional = true }
rustls-pki-types = { workspace = true, optional = true }
tracing = { workspace = true }
[lints]
workspace = true
[features]
default = ["ip"] # features that are enabled by default
ip = ["dep:local-ip-address"] # ip characteristics and their dependencies
tls = ["dep:rustls", "dep:rustls-pemfile", "dep:rustls-pki-types"] # tls characteristics and their dependencies
net = ["ip"] # empty network features
full = ["ip", "tls", "net"] # all features
+186
View File
@@ -0,0 +1,186 @@
use rustfs_config::{RUSTFS_TLS_CERT, RUSTFS_TLS_KEY};
use rustls::server::{ClientHello, ResolvesServerCert, ResolvesServerCertUsingSni};
use rustls::sign::CertifiedKey;
use rustls_pemfile::{certs, private_key};
use rustls_pki_types::{CertificateDer, PrivateKeyDer};
use std::collections::HashMap;
use std::io::Error;
use std::path::Path;
use std::sync::Arc;
use std::{fs, io};
use tracing::{debug, warn};
/// Load public certificate from file.
/// This function loads a public certificate from the specified file.
pub fn load_certs(filename: &str) -> io::Result<Vec<CertificateDer<'static>>> {
// Open certificate file.
let cert_file = fs::File::open(filename).map_err(|e| certs_error(format!("failed to open {}: {}", filename, e)))?;
let mut reader = io::BufReader::new(cert_file);
// Load and return certificate.
let certs = certs(&mut reader)
.collect::<Result<Vec<_>, _>>()
.map_err(|e| certs_error(format!("certificate file {} format error:{:?}", filename, e)))?;
if certs.is_empty() {
return Err(certs_error(format!(
"No valid certificate was found in the certificate file {}",
filename
)));
}
Ok(certs)
}
/// Load private key from file.
/// This function loads a private key from the specified file.
pub fn load_private_key(filename: &str) -> io::Result<PrivateKeyDer<'static>> {
// Open keyfile.
let keyfile = fs::File::open(filename).map_err(|e| certs_error(format!("failed to open {}: {}", filename, e)))?;
let mut reader = io::BufReader::new(keyfile);
// Load and return a single private key.
private_key(&mut reader)?.ok_or_else(|| certs_error(format!("no private key found in {}", filename)))
}
/// error function
pub fn certs_error(err: String) -> Error {
Error::new(io::ErrorKind::Other, err)
}
/// Load all certificates and private keys in the directory
/// This function loads all certificate and private key pairs from the specified directory.
/// It looks for files named `rustfs_cert.pem` and `rustfs_key.pem` in each subdirectory.
/// The root directory can also contain a default certificate/private key pair.
pub fn load_all_certs_from_directory(
dir_path: &str,
) -> io::Result<HashMap<String, (Vec<CertificateDer<'static>>, PrivateKeyDer<'static>)>> {
let mut cert_key_pairs = HashMap::new();
let dir = Path::new(dir_path);
if !dir.exists() || !dir.is_dir() {
return Err(certs_error(format!(
"The certificate directory does not exist or is not a directory: {}",
dir_path
)));
}
// 1. First check whether there is a certificate/private key pair in the root directory
let root_cert_path = dir.join(RUSTFS_TLS_CERT);
let root_key_path = dir.join(RUSTFS_TLS_KEY);
if root_cert_path.exists() && root_key_path.exists() {
debug!("find the root directory certificate: {:?}", root_cert_path);
let root_cert_str = root_cert_path
.to_str()
.ok_or_else(|| certs_error(format!("Invalid UTF-8 in root certificate path: {:?}", root_cert_path)))?;
let root_key_str = root_key_path
.to_str()
.ok_or_else(|| certs_error(format!("Invalid UTF-8 in root key path: {:?}", root_key_path)))?;
match load_cert_key_pair(root_cert_str, root_key_str) {
Ok((certs, key)) => {
// The root directory certificate is used as the default certificate and is stored using special keys.
cert_key_pairs.insert("default".to_string(), (certs, key));
}
Err(e) => {
warn!("unable to load root directory certificate: {}", e);
}
}
}
// 2.iterate through all folders in the directory
for entry in fs::read_dir(dir)? {
let entry = entry?;
let path = entry.path();
if path.is_dir() {
let domain_name = path
.file_name()
.and_then(|name| name.to_str())
.ok_or_else(|| certs_error(format!("invalid domain name directory:{:?}", path)))?;
// find certificate and private key files
let cert_path = path.join(RUSTFS_TLS_CERT); // e.g., rustfs_cert.pem
let key_path = path.join(RUSTFS_TLS_KEY); // e.g., rustfs_key.pem
if cert_path.exists() && key_path.exists() {
debug!("find the domain name certificate: {} in {:?}", domain_name, cert_path);
match load_cert_key_pair(cert_path.to_str().unwrap(), key_path.to_str().unwrap()) {
Ok((certs, key)) => {
cert_key_pairs.insert(domain_name.to_string(), (certs, key));
}
Err(e) => {
warn!("unable to load the certificate for {} domain name: {}", domain_name, e);
}
}
}
}
}
if cert_key_pairs.is_empty() {
return Err(certs_error(format!(
"No valid certificate/private key pair found in directory {}",
dir_path
)));
}
Ok(cert_key_pairs)
}
/// loading a single certificate private key pair
/// This function loads a certificate and private key from the specified paths.
/// It returns a tuple containing the certificate and private key.
fn load_cert_key_pair(cert_path: &str, key_path: &str) -> io::Result<(Vec<CertificateDer<'static>>, PrivateKeyDer<'static>)> {
let certs = load_certs(cert_path)?;
let key = load_private_key(key_path)?;
Ok((certs, key))
}
/// Create a multi-cert resolver
/// This function loads all certificates and private keys from the specified directory.
/// It uses the first certificate/private key pair found in the root directory as the default certificate.
/// The rest of the certificates/private keys are used for SNI resolution.
///
pub fn create_multi_cert_resolver(
cert_key_pairs: HashMap<String, (Vec<CertificateDer<'static>>, PrivateKeyDer<'static>)>,
) -> io::Result<impl ResolvesServerCert> {
#[derive(Debug)]
struct MultiCertResolver {
cert_resolver: ResolvesServerCertUsingSni,
default_cert: Option<Arc<CertifiedKey>>,
}
impl ResolvesServerCert for MultiCertResolver {
fn resolve(&self, client_hello: ClientHello) -> Option<Arc<CertifiedKey>> {
// try matching certificates with sni
if let Some(cert) = self.cert_resolver.resolve(client_hello) {
return Some(cert);
}
// If there is no matching SNI certificate, use the default certificate
self.default_cert.clone()
}
}
let mut resolver = ResolvesServerCertUsingSni::new();
let mut default_cert = None;
for (domain, (certs, key)) in cert_key_pairs {
// create a signature
let signing_key = rustls::crypto::aws_lc_rs::sign::any_supported_type(&key)
.map_err(|e| certs_error(format!("unsupported private key types:{}, err:{:?}", domain, e)))?;
// create a CertifiedKey
let certified_key = CertifiedKey::new(certs, signing_key);
if domain == "default" {
default_cert = Some(Arc::new(certified_key.clone()));
} else {
// add certificate to resolver
resolver
.add(&domain, certified_key)
.map_err(|e| certs_error(format!("failed to add a domain name certificate:{},err: {:?}", domain, e)))?;
}
}
Ok(MultiCertResolver {
cert_resolver: resolver,
default_cert,
})
}
@@ -1,4 +1,3 @@
use local_ip_address::{local_ip, local_ipv6};
use std::net::{IpAddr, Ipv4Addr};
/// Get the IP address of the machine
@@ -11,7 +10,9 @@ use std::net::{IpAddr, Ipv4Addr};
/// * `Some(IpAddr)` - Native IP address (IPv4 or IPv6)
/// * `None` - Unable to obtain any native IP address
pub fn get_local_ip() -> Option<IpAddr> {
local_ip().ok().or_else(|| local_ipv6().ok())
local_ip_address::local_ip()
.ok()
.or_else(|| local_ip_address::local_ipv6().ok())
}
/// Get the IP address of the machine as a string
+8
View File
@@ -0,0 +1,8 @@
mod certs;
mod ip;
mod net;
#[cfg(feature = "ip")]
pub use certs::*;
#[cfg(feature = "ip")]
pub use ip::*;
+1
View File
@@ -0,0 +1 @@
+28
View File
@@ -0,0 +1,28 @@
[package]
name = "zip"
edition.workspace = true
license.workspace = true
repository.workspace = true
rust-version.workspace = true
version.workspace = true
[dependencies]
async-compression = { version = "0.4.0", features = [
"tokio",
"bzip2",
"gzip",
"zlib",
"zstd",
"xz",
] }
# async_zip = { version = "0.0.17", features = ["tokio"] }
# rc-zip-tokio = "4.2.6"
tokio = { version = "1.45.0", features = ["full"] }
tokio-stream = "0.1.17"
tokio-tar = { workspace = true }
xz2 = { version = "0.1", optional = true, features = ["static"] }
[lints]
workspace = true
+124
View File
@@ -0,0 +1,124 @@
use async_compression::tokio::bufread::{BzDecoder, GzipDecoder, XzDecoder, ZlibDecoder, ZstdDecoder};
use tokio::io::{self, AsyncRead, BufReader};
use tokio_stream::StreamExt;
use tokio_tar::Archive;
#[derive(Debug, PartialEq)]
pub enum CompressionFormat {
Gzip, //.gz
Bzip2, //.bz2
// Lz4, //.lz4
Zip,
Xz, //.xz
Zlib, //.z
Zstd, //.zst
Unknown,
}
impl CompressionFormat {
pub fn from_extension(ext: &str) -> Self {
match ext {
"gz" => CompressionFormat::Gzip,
"bz2" => CompressionFormat::Bzip2,
// "lz4" => CompressionFormat::Lz4,
"zip" => CompressionFormat::Zip,
"xz" => CompressionFormat::Xz,
"zlib" => CompressionFormat::Zlib,
"zst" => CompressionFormat::Zstd,
_ => CompressionFormat::Unknown,
}
}
pub fn get_decoder<R>(&self, input: R) -> io::Result<Box<dyn AsyncRead + Send + Unpin>>
where
R: AsyncRead + Send + Unpin + 'static,
{
let reader = BufReader::new(input);
let decoder: Box<dyn AsyncRead + Send + Unpin + 'static> = match self {
CompressionFormat::Gzip => Box::new(GzipDecoder::new(reader)),
CompressionFormat::Bzip2 => Box::new(BzDecoder::new(reader)),
// CompressionFormat::Lz4 => Box::new(Lz4Decoder::new(reader)),
CompressionFormat::Zlib => Box::new(ZlibDecoder::new(reader)),
CompressionFormat::Xz => Box::new(XzDecoder::new(reader)),
CompressionFormat::Zstd => Box::new(ZstdDecoder::new(reader)),
_ => return Err(io::Error::new(io::ErrorKind::InvalidInput, "Unsupported file format")),
};
Ok(decoder)
}
}
pub async fn decompress<R, F>(input: R, format: CompressionFormat, mut callback: F) -> io::Result<()>
where
R: AsyncRead + Send + Unpin + 'static,
F: AsyncFnMut(tokio_tar::Entry<Archive<Box<dyn AsyncRead + Send + Unpin + 'static>>>) -> std::io::Result<()> + Send + 'static,
{
// 打开输入文件
// println!("format {:?}", format);
let decoder = format.get_decoder(input)?;
// let reader: BufReader<R> = BufReader::new(input);
// // 根据文件扩展名选择解压器
// let decoder: Box<dyn AsyncRead + Send + Unpin> = match format {
// CompressionFormat::Gzip => Box::new(GzipDecoder::new(reader)),
// CompressionFormat::Bzip2 => Box::new(BzDecoder::new(reader)),
// // CompressionFormat::Lz4 => Box::new(Lz4Decoder::new(reader)),
// CompressionFormat::Zlib => Box::new(ZlibDecoder::new(reader)),
// CompressionFormat::Xz => Box::new(XzDecoder::new(reader)),
// CompressionFormat::Zstd => Box::new(ZstdDecoder::new(reader)),
// // CompressionFormat::Zip => Box::new(DeflateDecoder::new(reader)),
// _ => {
// return Err(io::Error::new(io::ErrorKind::InvalidInput, "Unsupported file format"));
// }
// };
let mut ar = Archive::new(decoder);
let mut entries = ar.entries().unwrap();
while let Some(entry) = entries.next().await {
let f = match entry {
Ok(f) => f,
Err(e) => {
println!("Error reading entry: {}", e);
return Err(e);
}
};
// println!("{}", f.path().unwrap().display());
callback(f).await?;
}
Ok(())
}
// #[tokio::test]
// async fn test_decompress() -> io::Result<()> {
// use std::path::Path;
// use tokio::fs::File;
// let input_path = "/Users/weisd/Downloads/wsd.tar.gz"; // 替换为你的压缩文件路径
// let f = File::open(input_path).await?;
// let Some(ext) = Path::new(input_path).extension().and_then(|s| s.to_str()) else {
// return Err(io::Error::new(io::ErrorKind::InvalidInput, "Unsupported file format"));
// };
// match decompress(
// f,
// CompressionFormat::from_extension(ext),
// |entry: tokio_tar::Entry<Archive<Box<dyn AsyncRead + Send + Unpin>>>| async move {
// let path = entry.path().unwrap();
// println!("Extracted: {}", path.display());
// Ok(())
// },
// )
// .await
// {
// Ok(_) => println!("解压成功!"),
// Err(e) => println!("解压失败: {}", e),
// }
// Ok(())
// }