v1.0.0 发布合并回master (#52)

Co-authored-by: sundapeng <sundp@mail.zgclab.edu.cn>
Co-authored-by: xuxt <xuxt@zgclab.edu.cn>
Reviewed-on: #52

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/README.md

@@ -0,0 +1,181 @@
+# Fluent Bit 安装包
+
+这是一个 Fluent Bit 的自动化安装包，提供了完整的安装、卸载和健康检查功能。
+
+## 目录结构
+
+```
+fluent-bit-installer/
+├── install.sh                    # 安装脚本
+├── uninstall.sh                  # 卸载脚本
+├── package.sh                    # 打包脚本
+├── check_health.sh               # 健康检查脚本
+├── bin/
+│   └── fluent-bit_3.1.9_amd64.deb  # Fluent Bit 安装包
+└── config/
+    ├── fluent-bit.conf           # 主配置文件
+    ├── inject_labels.lua         # Lua 脚本
+    ├── parsers.conf              # 解析器配置
+    ├── inputs.d/                 # 输入配置目录
+    │   ├── 10-train.conf
+    │   └── 20-infer.conf
+    └── outputs.d/                # 输出配置目录
+        └── 10-es.conf
+```
+
+## 功能特性
+
+- **自动化安装**: 一键安装 Fluent Bit 及其依赖
+- **配置管理**: 自动部署预配置的配置文件
+- **服务管理**: 自动启动和停止 Fluent Bit 服务
+- **健康检查**: 提供 JSON 格式的健康状态检查
+- **完整卸载**: 彻底清理所有相关文件和配置
+- **用户管理**: 自动创建专用的 fluent-bit 用户
+
+## 使用方法
+
+### 1. 打包安装包
+
+```bash
+./package.sh
+```
+
+这将创建一个带时间戳的压缩包，例如：`fluent-bit-installer-20250924-160954.tar.gz`
+
+### 2. 安装 Fluent Bit
+
+```bash
+# 解压安装包
+tar -xzf fluent-bit-installer-*.tar.gz
+cd fluent-bit-installer-*
+
+# 运行安装脚本（需要 root 权限）
+sudo ./install.sh
+```
+
+### 3. 健康检查
+
+```bash
+./check_health.sh
+```
+
+输出示例：
+```json
+{"name": "fluent-bit", "status": "health", "reason": "success"}
+```
+
+### 4. 卸载 Fluent Bit
+
+```bash
+sudo ./uninstall.sh
+```
+
+## 安装后的文件位置
+
+- **二进制文件**: `/opt/fluent-bit/bin/fluent-bit`
+- **配置文件**: `/etc/fluent-bit/`
+- **日志文件**: `/var/log/fluent-bit/`
+- **缓冲区目录**: `/var/lib/fluent-bit/buffers/`
+- **运行用户**: `fluent-bit`
+- **HTTP 端口**: `2020`
+
+## 配置说明
+
+### 主配置文件
+
+主配置文件位于 `/etc/fluent-bit/fluent-bit.conf`，包含以下主要部分：
+
+- **SERVICE**: 服务配置，包括 HTTP 服务器设置
+- **INPUT**: 输入配置，通过 `inputs.d/` 目录管理
+- **FILTER**: 过滤器配置，包括解析器和标签注入
+- **OUTPUT**: 输出配置，通过 `outputs.d/` 目录管理
+
+### 输入配置
+
+- `10-train.conf`: 训练日志输入配置
+- `20-infer.conf`: 推理日志输入配置
+
+### 输出配置
+
+- `10-es.conf`: Elasticsearch 输出配置
+
+## 服务管理
+
+### 手动启动
+
+```bash
+/opt/fluent-bit/bin/fluent-bit --config=/etc/fluent-bit/fluent-bit.conf
+```
+
+### 后台启动
+
+```bash
+nohup /opt/fluent-bit/bin/fluent-bit --config=/etc/fluent-bit/fluent-bit.conf &
+```
+
+### 检查服务状态
+
+```bash
+# 检查进程
+ps aux | grep fluent-bit
+
+# 检查端口
+netstat -tuln | grep 2020
+
+# 检查日志
+tail -f /var/log/fluent-bit/fluent-bit.log
+```
+
+## API 接口
+
+Fluent Bit 提供 HTTP API 用于监控和管理：
+
+- **根路径**: `http://localhost:2020`
+- **状态接口**: `http://localhost:2020/api/v1/status`
+- **指标接口**: `http://localhost:2020/api/v1/metrics`
+
+## 故障排除
+
+### 常见问题
+
+1. **端口被占用**
+   - 检查端口 2020 是否被其他服务占用
+   - 修改配置文件中的端口设置
+
+2. **权限问题**
+   - 确保 fluent-bit 用户有足够的权限访问日志文件
+   - 检查目录权限设置
+
+3. **配置文件错误**
+   - 检查配置文件语法
+   - 查看日志文件中的错误信息
+
+### 日志查看
+
+```bash
+# 查看服务日志
+tail -f /var/log/fluent-bit/fluent-bit.log
+
+# 查看系统日志
+journalctl -u fluent-bit -f
+```
+
+## 系统要求
+
+- **操作系统**: Ubuntu/Debian/CentOS/RHEL/Fedora
+- **架构**: x86_64/amd64
+- **权限**: root 权限（用于安装和卸载）
+- **依赖**: curl（用于健康检查）
+
+## 版本信息
+
+- **Fluent Bit 版本**: 3.1.9
+- **安装包版本**: 1.0
+- **支持架构**: amd64
+
+## 注意事项
+
+1. 安装前请确保系统已更新
+2. 卸载时会保留 fluent-bit 用户（系统用户，可能被其他服务使用）
+3. 配置文件包含环境变量，请根据实际环境调整
+4. 建议在生产环境使用前进行充分测试

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/check_health.sh

@@ -0,0 +1,69 @@
+#!/bin/bash
+
+# Fluent Bit 健康检查脚本
+# 输出 JSON 格式结果
+
+set -e
+
+# 检查 Fluent Bit 健康状态
+check_health() {
+    local name="fluent-bit"
+    local status="unhealth"
+    local reason=""
+    local install_record="/opt/argus-metric/current/.install_record"
+    
+    # 首先尝试通过安装记录文件检查进程
+    if [[ -f "$install_record" ]]; then
+        # 尝试使用jq解析JSON格式的安装记录文件
+        local pid=""
+        if command -v jq &> /dev/null; then
+            pid=$(jq -r '.components."fluent-bit".pid // empty' "$install_record" 2>/dev/null || echo "")
+        else
+            # 如果没有jq，使用简单的文本解析方法
+            pid=$(grep -A 10 '"fluent-bit"' "$install_record" | grep '"pid"' | cut -d'"' -f4 | head -1)
+        fi
+        
+        if [[ -n "$pid" && "$pid" =~ ^[0-9]+$ ]]; then
+            if kill -0 "$pid" 2>/dev/null; then
+                # 进程存在且运行正常
+                status="health"
+                reason="进程运行正常 (PID: $pid)"
+                echo "{\"name\": \"$name\", \"status\": \"$status\", \"reason\": \"$reason\"}"
+                exit 0
+            else
+                reason="安装记录中的 PID $pid 进程不存在"
+                echo "{\"name\": \"$name\", \"status\": \"$status\", \"reason\": \"$reason\"}"
+                exit 1
+            fi
+        else
+            reason="安装记录文件中未找到有效的 fluent-bit PID"
+            echo "{\"name\": \"$name\", \"status\": \"$status\", \"reason\": \"$reason\"}"
+            exit 1
+        fi
+    else
+        # 如果安装记录文件不存在，尝试查找 fluent-bit 进程
+        local pids=$(pgrep -f "fluent-bit" 2>/dev/null || true)
+        if [[ -n "$pids" ]]; then
+            # 取第一个找到的 PID
+            local pid=$(echo "$pids" | head -1)
+            status="health"
+            reason="发现 fluent-bit 进程运行 (PID: $pid)，但未找到安装记录"
+            echo "{\"name\": \"$name\", \"status\": \"$status\", \"reason\": \"$reason\"}"
+            exit 0
+        else
+            reason="未找到 fluent-bit 进程，且安装记录文件不存在"
+            echo "{\"name\": \"$name\", \"status\": \"$status\", \"reason\": \"$reason\"}"
+            exit 1
+        fi
+    fi
+}
+
+# 主函数
+main() {
+    check_health
+}
+
+# 脚本入口
+if [[ "${BASH_SOURCE[0]}" == "${0}" ]]; then
+    main "$@"
+fi

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/config/fluent-bit.conf

@@ -0,0 +1,37 @@
+[SERVICE]
+    Daemon       Off
+    Parsers_File parsers.conf
+    HTTP_Server  On
+    HTTP_Listen  0.0.0.0
+    HTTP_Port    2020
+    storage.path /buffers
+    storage.sync normal
+    storage.checksum on
+    storage.backlog.mem_limit 128M
+    # 备注：该镜像默认未开启 Hot Reload，修改配置后请重启容器。
+
+@INCLUDE inputs.d/*.conf
+
+[FILTER]
+    Name   parser
+    Match  app.*
+    Key_Name log
+    Parser timestamp_parser
+    Reserve_Data On
+    Preserve_Key On
+    Unescape_Key On
+
+[FILTER]
+    Name   record_modifier
+    Match  *
+    Record cluster  ${CLUSTER}
+    Record rack     ${RACK}
+    Record host     ${HOSTNAME}
+
+[FILTER]
+    Name   lua
+    Match  app.*
+    script inject_labels.lua
+    call   add_labels
+
+@INCLUDE outputs.d/*.conf

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/config/inject_labels.lua

@@ -0,0 +1,15 @@
+function add_labels(tag, ts, record)
+  record["job_id"] = os.getenv("FB_JOB_ID") or record["job_id"] or "unknown"
+  record["user"]   = os.getenv("FB_USER")   or record["user"]   or "unknown"
+  record["model"]  = os.getenv("FB_MODEL")  or record["model"]  or "unknown"
+  record["gpu_id"] = os.getenv("FB_GPU_ID") or record["gpu_id"] or "na"
+  local p = record["log_path"] or ""
+  if string.find(p, "/logs/infer/") then
+    record["role"] = "infer"
+  elseif string.find(p, "/logs/train/") then
+    record["role"] = "train"
+  else
+    record["role"] = record["role"] or "app"
+  end
+  return 1, ts, record
+end

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/config/inputs.d/10-train.conf

@@ -0,0 +1,10 @@
+[INPUT]
+    Name              tail
+    Path              /logs/train/*.log
+    Tag               app.train
+    Path_Key          log_path
+    Refresh_Interval  5
+    DB                /buffers/train.db
+    Skip_Long_Lines   On
+    storage.type      filesystem
+    multiline.parser  python,go,java

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/config/inputs.d/20-infer.conf

@@ -0,0 +1,10 @@
+[INPUT]
+    Name              tail
+    Path              /logs/infer/*.log
+    Tag               app.infer
+    Path_Key          log_path
+    Refresh_Interval  5
+    DB                /buffers/infer.db
+    Skip_Long_Lines   On
+    storage.type      filesystem
+    multiline.parser  python,go,java

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/config/outputs.d/10-es.conf

@@ -0,0 +1,26 @@
+# 重要：使用 Logstash_Format + Logstash_Prefix，生成 train-*/infer-* 索引
+# 说明：Fluent Bit 配置仅支持 ${VAR} 占位符，不支持 Bash 的 ${VAR:-default}
+#      固定域名要求：使用 es.log.argus.com 与端口 9200
+[OUTPUT]
+    Name                es
+    Match               app.train
+    Host                es.log.argus.com
+    Port                9200
+    Logstash_Format     On
+    Logstash_Prefix     train
+    Replace_Dots        On
+    Generate_ID         On
+    Retry_Limit         False
+    Suppress_Type_Name  On
+
+[OUTPUT]
+    Name                es
+    Match               app.infer
+    Host                es.log.argus.com
+    Port                9200
+    Logstash_Format     On
+    Logstash_Prefix     infer
+    Replace_Dots        On
+    Generate_ID         On
+    Retry_Limit         False
+    Suppress_Type_Name  On

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/config/parsers.conf

@@ -0,0 +1,27 @@
+[MULTILINE_PARSER]
+    Name   python
+    Type   regex
+    Flush  2
+    Rule   "start_state"  "/^\d{4}-\d{2}-\d{2}[\sT]/"  "cont"
+    Rule   "cont"         "/^\s+|^Traceback|^\tat\s+/" "cont"
+
+[MULTILINE_PARSER]
+    Name   go
+    Type   regex
+    Flush  2
+    Rule   "start_state"  "/^[0-9]{4}\/[0-9]{2}\/[0-9]{2}/" "cont"
+    Rule   "cont"         "/^\s+|^\t/" "cont"
+
+[MULTILINE_PARSER]
+    Name   java
+    Type   regex
+    Flush  2
+    Rule   "start_state"  "/^\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}/" "cont"
+    Rule   "cont"         "/^\s+at\s+|^\t.../" "cont"
+
+[PARSER]
+    Name   timestamp_parser
+    Format regex
+    Regex  ^(?<timestamp>\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}(?:Z|[+-]\d{2}:?\d{2}))\s+(?<level>\w+)\s+(?<message>.*)$
+    Time_Key    timestamp
+    Time_Format %Y-%m-%dT%H:%M:%S%z

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/install.sh

@@ -0,0 +1,299 @@
+#!/bin/bash
+
+set -e
+
+# 颜色定义
+RED='\033[0;31m'
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+BLUE='\033[0;34m'
+NC='\033[0m' # No Color
+
+# 日志函数
+log_info() {
+    echo -e "${BLUE}[INFO]${NC} $1"
+}
+
+log_success() {
+    echo -e "${GREEN}[SUCCESS]${NC} $1"
+}
+
+log_warning() {
+    echo -e "${YELLOW}[WARNING]${NC} $1"
+}
+
+log_error() {
+    echo -e "${RED}[ERROR]${NC} $1"
+}
+
+log_info "Starting Fluent Bit installation..."
+
+# 解析命令行参数
+INSTALL_DIR="${1:-/opt/argus-metric/current}"
+
+# 更新安装记录
+update_install_record() {
+    local pid="$1"
+    # 使用传入的安装目录参数，如果没有则使用默认值
+    local install_base_dir="${2:-/opt/argus-metric/current}"
+    local install_record="$install_base_dir/.install_record"
+    
+    # 如果安装记录文件不存在，说明是首次安装，由主安装脚本统一创建
+    if [[ ! -f "$install_record" ]]; then
+        log_info "安装记录文件不存在，将由主安装脚本创建"
+        return 0
+    fi
+    
+    # 如果文件存在，说明是重启场景，只更新 PID 字段
+    if command -v jq &> /dev/null; then
+        # 读取当前 PID
+        local current_pid=$(jq -r '.components."fluent-bit".pid // ""' "$install_record" 2>/dev/null)
+        
+        if [[ -z "$current_pid" ]]; then
+            log_warning "无法读取当前 PID，跳过更新"
+            return 1
+        fi
+        
+        # 使用 jq 只更新 pid 字段，保持字符串类型，保留其他字段
+        jq --arg new_pid "$pid" '.components."fluent-bit".pid = $new_pid' "$install_record" > "$install_record.tmp" && mv "$install_record.tmp" "$install_record"
+        log_info "PID updated: $current_pid -> $pid"
+    else
+        log_warning "jq 命令不可用，无法更新安装记录文件"
+    fi
+}
+
+# 检查是否为 root 用户
+if [[ $EUID -ne 0 ]]; then
+    log_error "This script requires root privileges"
+    log_info "Please use: sudo $0"
+    exit 1
+fi
+
+# 停止可能运行的服务
+log_info "Stopping existing fluent-bit processes..."
+
+# 只匹配进程名为 fluent-bit 的进程
+pids=$(pgrep -x fluent-bit 2>/dev/null || true)
+
+if [[ -n "$pids" ]]; then
+    for pid in $pids; do
+        log_info "Stopping process PID: $pid"
+        kill "$pid" 2>/dev/null || true
+    done
+    sleep 2
+
+    # 检查是否还有残留进程
+    remaining_pids=$(pgrep -x fluent-bit 2>/dev/null || true)
+    if [[ -n "$remaining_pids" ]]; then
+        log_warning "Force killing unresponsive processes..."
+        for pid in $remaining_pids; do
+            kill -9 "$pid" 2>/dev/null || true
+        done
+    fi
+fi
+
+# 安装 Fluent Bit 依赖库 libpq5（离线模式）
+log_info "Checking Fluent Bit dependency: libpq5 ..."
+if ! ldconfig -p | grep -q libpq.so.5; then
+    if ls bin/libpq5_*.deb >/dev/null 2>&1; then
+        log_info "Installing local dependency package: libpq5"
+        DEBIAN_FRONTEND=noninteractive dpkg -i bin/libpq5_*.deb >/dev/null 2>&1 || {
+            log_error "Failed to install libpq5 from bin/, please check package validity"
+            exit 1
+        }
+    else
+        log_error "Missing dependency: libpq5 (libpq.so.5). Please put bin/libpq5_*.deb in the bin/ directory."
+        exit 1
+    fi
+else
+    log_info "libpq.so.5 already present on system"
+fi
+
+# 安装 Fluent Bit 依赖库 libyaml-0-2（离线模式）
+log_info "Checking Fluent Bit dependency: libyaml-0.so.2 ..."
+if ! ldconfig -p | grep -q libyaml-0.so.2; then
+    if ls bin/libyaml-0-2_*.deb >/dev/null 2>&1; then
+        log_info "Installing local dependency package: libyaml-0-2"
+        DEBIAN_FRONTEND=noninteractive dpkg -i bin/libyaml-0-2_*.deb >/dev/null 2>&1 || {
+            log_error "Failed to install libyaml-0-2 from bin/, please check package validity"
+            exit 1
+        }
+    else
+        log_error "Missing dependency: libyaml-0-2 (libyaml-0.so.2). Please put bin/libyaml-0-2_*.deb in the bin/ directory."
+        exit 1
+    fi
+else
+    log_info "libyaml-0.so.2 already present on system"
+fi
+
+# 清理可能存在的旧 fluent-bit 安装（避免配置文件冲突）
+log_info "Cleaning up old fluent-bit installation if exists..."
+if dpkg -l | grep -q "^ii.*fluent-bit"; then
+    log_info "Found existing fluent-bit package, removing..."
+    dpkg --purge fluent-bit 2>/dev/null || true
+    apt-get remove --purge -y fluent-bit 2>/dev/null || true
+fi
+
+# 确保清理残留的配置文件
+if [[ -d "/etc/fluent-bit" ]]; then
+    log_info "Removing old fluent-bit configuration directory..."
+    rm -rf /etc/fluent-bit
+fi
+
+# 安装 Fluent Bit 主包
+log_info "Installing Fluent Bit from deb package..."
+deb_file="bin/fluent-bit_3.1.9_amd64.deb"
+if [[ ! -f "$deb_file" ]]; then
+    log_error "Fluent Bit package not found: $deb_file"
+    exit 1
+fi
+
+DEBIAN_FRONTEND=noninteractive dpkg -i "$deb_file" >/dev/null 2>&1 || true
+
+# 验证 Fluent Bit 可以运行
+fb_version=$(/opt/fluent-bit/bin/fluent-bit --version 2>&1 | head -1)
+log_info "Fluent Bit version: $fb_version"
+
+# 创建 fluent-bit 用户
+log_info "Creating fluent-bit user..."
+if ! id "fluent-bit" &>/dev/null; then
+    useradd --no-create-home --shell /bin/false fluent-bit
+fi
+
+# 创建配置目录
+log_info "Installing configuration files..."
+mkdir -p /etc/fluent-bit
+if [[ -d "config" ]]; then
+    cp -r config/* /etc/fluent-bit/
+    chown -R fluent-bit:fluent-bit /etc/fluent-bit
+fi
+
+# 创建日志和缓冲区目录
+log_info "Creating log and buffer directories..."
+mkdir -p /logs/train /logs/infer /buffers
+# 对共享日志目录采用 1777（含粘滞位），便于宿主任意账号创建文件/目录
+if [[ "${ARGUS_LOGS_WORLD_WRITABLE:-1}" == "1" ]]; then
+    chmod 1777 /logs/train /logs/infer || true
+else
+    chmod 755 /logs/train /logs/infer || true
+fi
+# 缓冲目录限进程使用
+chmod 770 /buffers || true
+# 目录属主设置，不影响 1777 粘滞位
+chown -R fluent-bit:fluent-bit /logs /buffers 2>/dev/null || true
+
+# 启动 Fluent Bit
+log_info "Starting Fluent Bit with configuration from /etc/fluent-bit/"
+config_path="/etc/fluent-bit/fluent-bit.conf"
+
+if [[ ! -f "$config_path" ]]; then
+    log_error "Configuration file not found: $config_path"
+    exit 1
+fi
+
+# 设置环境变量
+log_info "Setting environment variables..."
+
+# 获取非 127.0.0.1 的 IP 地址作为 HOSTNAME
+if [[ -z "${HOSTNAME:-}" ]]; then
+    # 获取 177.x.x.x 段的 IP 地址
+    HOSTNAME=$(ip route get 8.8.8.8 2>/dev/null | grep -oP 'src \K\S+' | grep '^177\.' | head -1)
+    
+    # 如果没有找到 177.x.x.x 段的 IP，则获取第一个非 127.0.0.1 的 IP
+    if [[ -z "$HOSTNAME" ]]; then
+        HOSTNAME=$(ip route get 8.8.8.8 2>/dev/null | grep -oP 'src \K\S+' | grep -v '^127\.' | head -1)
+    fi
+    
+    # 如果还是没有找到，使用 hostname 命令
+    if [[ -z "$HOSTNAME" ]]; then
+        HOSTNAME=$(hostname)
+    fi
+fi
+export HOSTNAME
+
+export CLUSTER="${CLUSTER:-local}"
+export RACK="${RACK:-dev}"
+# 默认使用固定域名（满足“固定域名”需求）；若外部传入覆盖，则使用外部值
+export ES_HOST="${ES_HOST:-es.log.argus.com}"
+export ES_PORT="${ES_PORT:-9200}"
+
+log_info "Environment variables:"
+log_info "  CLUSTER=$CLUSTER"
+log_info "  RACK=$RACK"
+log_info "  HOSTNAME=$HOSTNAME"
+log_info "  ES_HOST=$ES_HOST"
+log_info "  ES_PORT=$ES_PORT"
+
+# 检查 fluent-bit 二进制文件
+log_info "[DEBUG] Checking fluent-bit binary..."
+if [[ ! -f "/opt/fluent-bit/bin/fluent-bit" ]]; then
+    log_error "fluent-bit binary not found at /opt/fluent-bit/bin/fluent-bit"
+    exit 1
+fi
+log_info "[DEBUG] fluent-bit binary exists and is executable: $(ls -lh /opt/fluent-bit/bin/fluent-bit)"
+
+# 检查配置文件
+log_info "[DEBUG] Checking configuration file: $config_path"
+if [[ ! -f "$config_path" ]]; then
+    log_error "Configuration file not found: $config_path"
+    exit 1
+fi
+log_info "[DEBUG] Configuration file exists: $(ls -lh $config_path)"
+
+# 显示完整的启动命令
+log_info "[DEBUG] Full command to execute:"
+log_info "[DEBUG]   su -s /bin/bash fluent-bit -c 'env CLUSTER=\"$CLUSTER\" RACK=\"$RACK\" HOSTNAME=\"$HOSTNAME\" ES_HOST=\"$ES_HOST\" ES_PORT=\"$ES_PORT\" /opt/fluent-bit/bin/fluent-bit --config=\"$config_path\"'"
+
+# 清空或创建日志文件
+log_info "[DEBUG] Preparing log file: /var/log/fluent-bit.log"
+: > /var/log/fluent-bit.log
+chmod 666 /var/log/fluent-bit.log
+
+log_info "Command: /opt/fluent-bit/bin/fluent-bit --config=$config_path"
+log_info "[DEBUG] Starting fluent-bit process as fluent-bit user (using su)..."
+nohup su -s /bin/bash fluent-bit -c "env CLUSTER='$CLUSTER' RACK='$RACK' HOSTNAME='$HOSTNAME' ES_HOST='$ES_HOST' ES_PORT='$ES_PORT' /opt/fluent-bit/bin/fluent-bit --config='$config_path' >> /var/log/fluent-bit.log 2>&1" &
+
+bg_pid=$!
+log_info "[DEBUG] Background process started with PID: $bg_pid"
+
+# 等待服务启动
+log_info "[DEBUG] Waiting 3 seconds for service to start..."
+sleep 3
+
+# 查找实际的 fluent-bit 进程 PID
+log_info "[DEBUG] Searching for fluent-bit process..."
+log_info "[DEBUG] Running: pgrep -u fluent-bit -x fluent-bit"
+actual_pid=$(pgrep -u fluent-bit -x fluent-bit | head -1)
+
+# 显示所有 fluent-bit 相关进程
+log_info "[DEBUG] All fluent-bit related processes:"
+ps aux | grep fluent-bit | grep -v grep || log_warning "No fluent-bit processes found in ps output"
+
+if [[ -n "$actual_pid" ]]; then
+    log_success "Fluent Bit started successfully (PID: $actual_pid)"
+    log_info "[DEBUG] Process details: $(ps -p $actual_pid -o pid,user,cmd --no-headers)"
+    
+    # 更新安装记录
+    update_install_record "$actual_pid" "$INSTALL_DIR"
+else
+    log_error "Fluent Bit failed to start - no fluent-bit process found"
+    log_info "[DEBUG] Checking if background process $bg_pid still exists..."
+    if ps -p $bg_pid > /dev/null 2>&1; then
+        log_warning "Background shell process $bg_pid still exists"
+    else
+        log_warning "Background shell process $bg_pid has exited"
+    fi
+    
+    log_info "[DEBUG] Last 20 lines of /var/log/fluent-bit.log:"
+    if [[ -f "/var/log/fluent-bit.log" ]]; then
+        tail -20 /var/log/fluent-bit.log | while IFS= read -r line; do
+            log_info "[LOG] $line"
+        done
+    else
+        log_error "Log file /var/log/fluent-bit.log does not exist"
+    fi
+    
+    exit 1
+fi
+
+log_success "Fluent Bit installation completed!"

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/package.sh

@@ -0,0 +1,87 @@
+#!/bin/bash
+
+set -e
+
+# 颜色定义
+GREEN='\033[0;32m'
+BLUE='\033[0;34m'
+NC='\033[0m' # No Color
+
+log_info() {
+    echo -e "${BLUE}[INFO]${NC} $1"
+}
+
+log_success() {
+    echo -e "${GREEN}[SUCCESS]${NC} $1"
+}
+
+# 获取当前目录
+CURRENT_DIR=$(pwd)
+PACKAGE_NAME="fluent-bit-$(date +%Y%m%d-%H%M%S)"
+PACKAGE_FILE="${PACKAGE_NAME}.tar.gz"
+
+log_info "开始打包 Fluent Bit 安装包..."
+
+# 检查必要文件
+log_info "检查必要文件..."
+
+required_files=(
+    "install.sh"
+    "uninstall.sh"
+    "bin/fluent-bit_3.1.9_amd64.deb"
+    "check_health.sh"
+)
+
+missing_files=()
+for file in "${required_files[@]}"; do
+    if [[ ! -f "$file" ]]; then
+        missing_files+=("$file")
+    fi
+done
+
+if [[ ${#missing_files[@]} -gt 0 ]]; then
+    echo "缺少以下文件:"
+    for file in "${missing_files[@]}"; do
+        echo "  - $file"
+    done
+    exit 1
+fi
+
+log_success "所有必要文件检查完成"
+
+# 创建临时目录
+TEMP_DIR=$(mktemp -d)
+log_info "创建临时目录: $TEMP_DIR"
+
+# 复制文件到临时目录
+cp -r . "$TEMP_DIR/$PACKAGE_NAME"
+
+# 进入临时目录
+cd "$TEMP_DIR"
+
+# 创建压缩包
+log_info "创建压缩包: $PACKAGE_FILE"
+tar -czf "$PACKAGE_FILE" "$PACKAGE_NAME"
+
+# 移动压缩包到原目录
+mv "$PACKAGE_FILE" "$CURRENT_DIR/"
+
+# 清理临时目录
+rm -rf "$TEMP_DIR"
+
+# 返回原目录
+cd "$CURRENT_DIR"
+
+# 显示结果
+log_success "打包完成！"
+echo
+echo "安装包文件: $PACKAGE_FILE"
+echo "文件大小: $(du -h "$PACKAGE_FILE" | cut -f1)"
+echo
+echo "使用方法:"
+echo "1. 将 $PACKAGE_FILE 传输到目标服务器"
+echo "2. 解压: tar -xzf $PACKAGE_FILE"
+echo "3. 进入目录: cd $PACKAGE_NAME"
+echo "4. 运行安装: sudo ./install.sh"
+echo
+echo "注意: 请确保所有必要文件都存在"

src/metric/client-plugins/all-in-one-full/plugins/fluent-bit/uninstall.sh

@@ -0,0 +1,169 @@
+#!/bin/bash
+set -euo pipefail
+
+echo "[INFO] Starting Fluent Bit uninstallation..."
+
+# 检查是否为 root 用户
+if [[ $EUID -ne 0 ]]; then
+    echo "[ERROR] This script requires root privileges"
+    echo "[INFO] Please use: sudo $0"
+    exit 1
+fi
+
+echo "[WARNING] This operation will completely uninstall Fluent Bit"
+read -p "Confirm to continue? (y/N): " confirm
+
+if [[ "$confirm" != "y" && "$confirm" != "Y" ]]; then
+    echo "[INFO] Uninstallation cancelled"
+    exit 0
+fi
+
+# 停止运行中的进程
+echo "[INFO] Stopping Fluent Bit processes..."
+install_record="/opt/argus-metric/current/.install_record"
+stopped=false
+
+# 首先尝试通过安装记录文件停止服务
+if [[ -f "$install_record" ]]; then
+    # 尝试使用jq解析JSON格式的安装记录文件
+    pid=""
+    if command -v jq &> /dev/null; then
+        pid=$(jq -r '.components."fluent-bit".pid // empty' "$install_record" 2>/dev/null || echo "")
+    else
+        # 如果没有jq，使用简单的文本解析方法
+        pid=$(grep -A 10 '"fluent-bit"' "$install_record" | grep '"pid"' | cut -d'"' -f4 | head -1)
+    fi
+    
+    if [[ -n "$pid" && "$pid" =~ ^[0-9]+$ ]]; then
+        if kill -0 "$pid" 2>/dev/null; then
+            echo "[INFO] Stopping service via installation record (PID: $pid)..."
+            kill "$pid"
+            sleep 3
+            
+            # 检查进程是否已停止
+            if kill -0 "$pid" 2>/dev/null; then
+                echo "[WARNING] Process unresponsive, force killing..."
+                kill -9 "$pid" 2>/dev/null || true
+            fi
+            echo "[SUCCESS] Fluent Bit process stopped"
+            stopped=true
+        else
+            echo "[WARNING] PID in installation record no longer exists"
+        fi
+    fi
+fi
+
+# 查找并杀死所有 fluent-bit 进程
+pids=$(pgrep -f "fluent-bit" 2>/dev/null || true)
+if [[ -n "$pids" ]]; then
+    echo "[INFO] Found fluent-bit processes, stopping..."
+    for pid in $pids; do
+        echo "[INFO] Stopping process PID: $pid"
+        kill "$pid" 2>/dev/null || true
+    done
+    sleep 2
+    
+    # 检查是否还有进程在运行，如果有则强制终止
+    remaining_pids=$(pgrep -f "fluent-bit" 2>/dev/null || true)
+    if [[ -n "$remaining_pids" ]]; then
+        echo "[WARNING] Processes unresponsive, force killing..."
+        for pid in $remaining_pids; do
+            echo "[INFO] Force killing process PID: $pid"
+            kill -9 "$pid" 2>/dev/null || true
+        done
+        sleep 1
+    fi
+    
+    # 最终检查
+    if pgrep -f "fluent-bit" > /dev/null; then
+        echo "[ERROR] Unable to stop all fluent-bit processes"
+    else
+        echo "[SUCCESS] All Fluent Bit processes stopped"
+        stopped=true
+    fi
+else
+    echo "[INFO] No Fluent Bit processes running"
+fi
+
+if [[ "$stopped" == "false" ]]; then
+    echo "[WARNING] No Fluent Bit processes found to stop"
+fi
+
+# 卸载 Fluent Bit 包
+echo "[INFO] Uninstalling Fluent Bit package..."
+if dpkg -l | grep -q "fluent-bit"; then
+    echo "[INFO] Found fluent-bit package installed via dpkg, uninstalling..."
+    dpkg --remove --force-remove-reinstreq fluent-bit || true
+    echo "[SUCCESS] Fluent Bit package uninstalled"
+else
+    echo "[INFO] No fluent-bit package found via package manager"
+fi
+
+# 删除二进制文件
+echo "[INFO] Removing Fluent Bit binary files..."
+binary_dir="/opt/fluent-bit"
+if [[ -d "$binary_dir" ]]; then
+    rm -rf "$binary_dir"
+    echo "[SUCCESS] Binary directory removed: $binary_dir"
+else
+    echo "[INFO] Binary directory does not exist"
+fi
+
+# 删除配置文件
+echo "[INFO] Removing configuration files..."
+config_dir="/etc/fluent-bit"
+if [[ -d "$config_dir" ]]; then
+    rm -rf "$config_dir"
+    echo "[SUCCESS] Configuration directory removed"
+else
+    echo "[INFO] Configuration directory does not exist"
+fi
+
+# 删除数据目录
+echo "[INFO] Removing data directories..."
+data_dirs=("/logs" "/buffers")
+deleted=false
+for data_dir in "${data_dirs[@]}"; do
+    if [[ -d "$data_dir" ]]; then
+        rm -rf "$data_dir"
+        echo "[SUCCESS] Data directory removed: $data_dir"
+        deleted=true
+    fi
+done
+
+if [[ "$deleted" == "false" ]]; then
+    echo "[INFO] No data directories found"
+fi
+
+# 清理安装记录
+echo "[INFO] Cleaning up installation record..."
+if [[ -f "$install_record" ]]; then
+    # 从安装记录中移除 fluent-bit 条目
+    sed -i '/^fluent-bit:/d' "$install_record"
+    echo "[SUCCESS] Installation record cleaned"
+else
+    echo "[INFO] Installation record file does not exist"
+fi
+
+# 检查用户状态
+echo "[INFO] Checking fluent-bit user status..."
+if id "fluent-bit" &>/dev/null; then
+    echo "[INFO] fluent-bit user exists"
+    echo "[WARNING] fluent-bit is a system user, may be used by other services"
+    echo "[INFO] fluent-bit user will be preserved for system stability"
+    echo "[INFO] To manually remove, run: sudo userdel fluent-bit"
+else
+    echo "[INFO] fluent-bit user does not exist"
+fi
+
+echo "[SUCCESS] Fluent Bit uninstallation completed!"
+echo
+echo "Removed content:"
+echo "  - Binary directory: /opt/fluent-bit"
+echo "  - Configuration directory: /etc/fluent-bit"
+echo "  - Application log directory: /logs"
+echo "  - Buffer directory: /buffers"
+echo
+echo "Note:"
+echo "  - fluent-bit user preserved (system user, may be used by other services)"
+echo "  - For complete cleanup, manually check and remove related files"