docs: 更新 README.md 配置示例和文档

- 配置示例:server -> servers 数组格式,lua phases -> scripts 格式
- 负载均衡表:添加 Random 算法(Power of Two Choices)
- 依赖表:补充 fasthttp/router, brotli, golang-lru, uuid 等 9 个遗漏依赖
- Lua API:补充 ngx.req/resp/timer/location.ctx/balancer 详细 API
- 目录结构:补充 http2/, matcher/, converter/, mimeutil/ 等缺失模块
- 状态端点:重写 JSON 示例以匹配实际 status.go 输出结构
- Go 版本:1.26+ -> 1.24+(当前最新稳定版)

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This commit is contained in:
xfy 2026-04-24 23:51:38 +08:00
parent 2122067efb
commit 13a2a3d8b2

714
README.md
View File

@ -35,6 +35,7 @@
| Least Connections | 最少连接,选择活跃连接最少的目标 |
| IP Hash | IP 哈希,同一客户端始终路由到同一目标 |
| Consistent Hash | 一致性哈希,支持虚拟节点(默认 150最小化节点变更影响 |
| Random | Power of Two Choices随机选择两个后比较简单高效 |
### 故障转移
@ -83,7 +84,7 @@ proxy:
```bash
# 克隆仓库
git clone https://github.com/xfy/lolly.git
git clone https://github.com/DefectingCat/lolly.git
cd lolly
# 本地构建
@ -149,129 +150,283 @@ lolly 支持将 nginx 配置文件转换为 YAML 格式:
配置文件使用 YAML 格式。以下是完整配置示例:
```yaml
server:
listen: ":8080"
name: "example.com"
read_timeout: 30s
write_timeout: 30s
idle_timeout: 120s
max_conns_per_ip: 100
max_requests_per_conn: 1000
client_max_body_size: "10MB"
mode: "auto" # single/vhost/multi_server/auto
static:
- path: "/"
root: "/var/www/html"
index: ["index.html", "index.htm"]
try_files: ["$uri", "$uri/", "/index.html"]
try_files_pass: false
- path: "/assets/"
root: "/var/www/assets"
variables:
set:
app_name: "lolly"
version: "1.0.0"
proxy:
- path: "/api"
targets:
- url: "http://backend1:8080"
weight: 2
- url: "http://backend2:8080"
weight: 1
load_balance: "weighted_round_robin"
health_check:
interval: 10s
path: "/health"
timeout: 5s
next_upstream:
tries: 3
http_codes: [502, 503, 504]
timeout:
connect: 5s
read: 30s
write: 30s
headers:
set_request:
X-Forwarded-For: "$remote_addr"
X-Real-IP: "$remote_addr"
set_response:
X-Proxy-By: "lolly"
cache:
servers:
- listen: ":8080"
name: "example.com"
default: true # vhost 默认主机
server_names: ["example.com", "*.example.com"]
read_timeout: 30s
write_timeout: 30s
idle_timeout: 120s
max_conns_per_ip: 100
max_requests_per_conn: 1000
concurrency: 262144 # 最大并发连接数
read_buffer_size: 16384 # 16KB
write_buffer_size: 16384
client_max_body_size: "10MB"
server_tokens: false # 隐藏版本号
static:
- path: "/"
root: "/var/www/html"
index: ["index.html", "index.htm"]
try_files: ["$uri", "$uri/", "/index.html"]
try_files_pass: false
symlink_check: true
location_type: "prefix"
- path: "/assets/"
root: "/var/www/assets"
internal: true # 仅允许内部重定向访问
proxy:
- path: "/api"
targets:
- url: "http://backend1:8080"
weight: 2
max_conns: 100
max_fails: 3
fail_timeout: 30s
backup: false
down: false
proxy_uri: "/v1" # 替换请求路径
- url: "http://backend2:8080"
weight: 1
load_balance: "weighted_round_robin"
virtual_nodes: 150 # 一致性哈希虚拟节点数
hash_key: "$uri" # 一致性哈希键
location_type: "prefix_priority"
internal: false
health_check:
interval: 10s
path: "/health"
timeout: 5s
slow_start: 30s # 慢启动时间
match:
status: ["200-299"]
body: "ok"
headers:
Content-Type: "application/json"
next_upstream:
tries: 3
http_codes: [502, 503, 504]
timeout:
connect: 5s
read: 30s
write: 30s
headers:
set_request:
X-Forwarded-For: "$remote_addr"
X-Real-IP: "$remote_addr"
set_response:
X-Proxy-By: "lolly"
remove: ["X-Internal-Header"]
hide_response: ["X-Powered-By"]
pass_response: ["Content-Type", "Content-Length"]
cookie_domain: ".example.com"
cookie_path: "/"
cache:
enabled: true
max_age: 5m
stale_while_revalidate: 1m
stale_if_error: 10m
stale_if_timeout: 5m
cache_lock: true
cache_lock_timeout: 5s
methods: ["GET", "HEAD"]
min_uses: 1 # 缓存阈值
background_update_disable: false
cache_ignore_headers: ["Set-Cookie"]
revalidate: true # 条件请求
cache_valid:
ok: 10m # 200-299
redirect: 1h # 301/302
not_found: 1m
client_error: 0
server_error: 0
buffering:
mode: "default" # default/on/off
buffer_size: 16384
buffers: "8 16k" # 8 个 16KB 缓冲区
proxy_ssl:
enabled: true
server_name: "backend.internal"
trusted_ca: "/etc/ssl/ca/upstream-ca.crt"
client_cert: "/etc/ssl/client.crt"
client_key: "/etc/ssl/client.key"
min_version: "TLSv1.2"
insecure_skip_verify: false
redirect_rewrite:
mode: "default" # default/off/custom
rules:
- pattern: "http://backend:8000/"
replacement: "$scheme://$host:$server_port/"
client_max_body_size: "50MB"
balancer_by_lua:
enabled: false
script: "/etc/lolly/scripts/balancer.lua"
timeout: 100ms
fallback: "round_robin"
ssl:
cert: "/etc/ssl/certs/server.crt"
key: "/etc/ssl/private/server.key"
cert_chain: "/etc/ssl/certs/chain.crt"
protocols: ["TLSv1.2", "TLSv1.3"]
ciphers: ["TLS_AES_128_GCM_SHA256", "TLS_AES_256_GCM_SHA384"]
ocsp_stapling: true
session_tickets:
enabled: true
max_age: 5m
cache_lock: true
stale_while_revalidate: 1m
client_max_body_size: "50MB"
key_file: "/var/lib/lolly/session_tickets.key"
rotate_interval: 1h
retain_keys: 3
client_verify:
enabled: false
mode: "require" # none/request/require/optional_no_ca
client_ca: "/etc/ssl/ca/client-ca.crt"
verify_depth: 1
crl: "/etc/ssl/ca/client-ca.crl"
http2:
enabled: true
max_concurrent_streams: 128
max_header_list_size: 16384
idle_timeout: 90s
push_enabled: false
h2c_enabled: false
graceful_shutdown_timeout: 30s
hsts:
max_age: 31536000
include_sub_domains: true
preload: false
ssl:
cert: "/etc/ssl/certs/server.crt"
key: "/etc/ssl/private/server.key"
cert_chain: "/etc/ssl/certs/chain.crt"
protocols: ["TLSv1.2", "TLSv1.3"]
ocsp_stapling: true
hsts:
max_age: 31536000
include_sub_domains: true
preload: false
security:
access:
allow: ["192.168.1.0/24", "10.0.0.0/8"]
deny: []
default: "deny"
trusted_proxies: ["172.16.0.0/16"]
geoip:
enabled: false
database: "/var/lib/GeoIP/GeoIP2-Country.mmdb"
allow_countries: ["CN", "US"]
deny_countries: []
default: "deny"
cache_size: 10000
cache_ttl: 1h
private_ip_behavior: "allow"
rate_limit:
request_rate: 100
burst: 200
conn_limit: 50
algorithm: "token_bucket" # token_bucket/sliding_window
key: "ip"
sliding_window_mode: "approximate" # exact/approximate
sliding_window: 60
auth:
type: "basic"
require_tls: true
algorithm: "bcrypt" # bcrypt/argon2id
realm: "Secure Area"
min_password_length: 8
users:
- name: "admin"
password: "$2y$10$N9qo8uLOickgx2ZMRZoMy..."
auth_request:
enabled: false
uri: "/auth"
method: "GET"
auth_timeout: 5s
forward_headers: ["Cookie", "Authorization"]
headers:
X-Original-Uri: "$request_uri"
X-Original-Host: "$host"
headers:
x_frame_options: "DENY"
x_content_type_options: "nosniff"
content_security_policy: "default-src 'self'"
referrer_policy: "strict-origin-when-cross-origin"
permissions_policy: "geolocation=(), microphone=()"
error_page:
pages:
404: "/var/www/errors/404.html"
500: "/var/www/errors/500.html"
default: "/var/www/errors/error.html"
response_code: 0 # 覆盖响应状态码
security:
access:
allow: ["192.168.1.0/24", "10.0.0.0/8"]
deny: []
default: "deny"
trusted_proxies: ["172.16.0.0/16"]
geoip:
database: "/var/lib/GeoIP/GeoIP2-Country.mmdb"
allow_countries: ["CN", "US"]
deny_countries: []
default: "deny"
rate_limit:
request_rate: 100
burst: 200
conn_limit: 50
algorithm: "token_bucket"
auth:
type: "basic"
require_tls: true
algorithm: "bcrypt"
realm: "Secure Area"
users:
- name: "admin"
password: "$2y$10$N9qo8uLOickgx2ZMRZoMy..."
headers:
x_frame_options: "DENY"
x_content_type_options: "nosniff"
content_security_policy: "default-src 'self'"
referrer_policy: "strict-origin-when-cross-origin"
error_page:
pages:
404: "/var/www/errors/404.html"
500: "/var/www/errors/500.html"
default: "/var/www/errors/error.html"
rewrite:
- pattern: "^/old/(.*)$"
replacement: "/new/$1"
flag: "permanent"
- pattern: "^/api/v1/(.*)$"
replacement: "/v1/$1"
flag: "last"
rewrite:
- pattern: "^/old/(.*)$"
replacement: "/new/$1"
flag: "permanent"
compression:
type: "gzip" # gzip/brotli/both
level: 6
min_size: 1024
types: ["text/html", "text/css", "application/javascript", "application/json"]
gzip_static: true
gzip_static_extensions: [".gz", ".br"]
compression:
type: "gzip"
level: 6
min_size: 1024
types: ["text/html", "text/css", "application/javascript", "application/json"]
gzip_static: true
gzip_static_extensions: [".gz", ".br"]
lua:
enabled: true
package_path: "/etc/lolly/lua/?.lua"
code_cache:
lua:
enabled: true
ttl: 60s
max_concurrent: 1000
timeout: 30s
phases:
rewrite: "/etc/lolly/lua/rewrite.lua"
access: "/etc/lolly/lua/access.lua"
content: "/etc/lolly/lua/content.lua"
log: "/etc/lolly/lua/log.lua"
scripts:
- path: "/etc/lolly/lua/rewrite.lua"
phase: "rewrite"
timeout: 30s
enabled: true
- path: "/etc/lolly/lua/access.lua"
phase: "access"
timeout: 30s
- path: "/etc/lolly/lua/content.lua"
phase: "content"
timeout: 30s
- path: "/etc/lolly/lua/log.lua"
phase: "log"
timeout: 30s
global_settings:
max_concurrent_coroutines: 1000
coroutine_timeout: 30s
code_cache_size: 1000
max_execution_time: 30s
coroutine_stack_size: 64
coroutine_pool_warmup: 4
enable_file_watch: true
minimize_stack_memory: true
limit_rate:
rate: 1048576 # 1MB/s
burst: 524288 # 512KB 突发
large_file_threshold: 10485760 # 10MB
large_file_strategy: "skip" # skip/coarse
types:
default_type: "application/octet-stream"
map:
".html": "text/html"
".css": "text/css"
".js": "application/javascript"
".json": "application/json"
unix_socket:
mode: 0666
user: "www-data"
group: "www-data"
cache_api:
enabled: false
path: "/_cache/purge"
allow: ["127.0.0.1", "10.0.0.0/8"]
auth:
type: "token"
token: "${CACHE_API_TOKEN}"
http3:
enabled: true
@ -290,15 +445,54 @@ stream:
- addr: "mysql2:3306"
weight: 1
load_balance: "round_robin"
ssl:
enabled: false
cert: "/etc/ssl/stream.crt"
key: "/etc/ssl/stream.key"
client_ca: ""
protocols: ["TLSv1.2", "TLSv1.3"]
ciphers: []
verify_depth: 1
proxy_ssl:
enabled: false
trusted_ca: "/etc/ssl/ca.crt"
server_name: "mysql.internal"
cert: ""
key: ""
protocols: ["TLSv1.2", "TLSv1.3"]
verify: true
session_reuse: true
resolver:
enabled: false
addresses:
- "8.8.8.8:53"
- "8.8.4.4:53"
valid: 30s # 缓存 TTL
timeout: 5s
ipv4: true
ipv6: false
cache_size: 1024
cache_path:
path: "/var/cache/lolly"
levels: "1:2"
max_size: 1073741824 # 1GB
inactive: 60m
purger: true
purger_interval: 1m
l1_max_entries: 10000
l1_max_size: 67108864 # 64MB
promote_threshold: 3
logging:
format: "json"
format: "json" # text/json
access:
path: "/var/log/lolly/access.log"
format: "combined"
error:
path: "/var/log/lolly/error.log"
level: "info"
level: "info" # debug/info/warn/error
performance:
goroutine_pool:
@ -316,12 +510,21 @@ performance:
monitoring:
status:
enabled: true
path: "/status"
format: "json" # json/text/html/prometheus
allow: ["127.0.0.1", "10.0.0.0/8"]
pprof:
enabled: false
path: "/debug/pprof"
allow: ["127.0.0.1"]
shutdown:
graceful_timeout: 30s # SIGQUIT 优雅停止
fast_timeout: 5s # SIGINT/SIGTERM 快速停止
include:
- path: "conf.d/*.yaml" # glob 模式展开
```
完整配置说明请参考 `config.example.yaml` 和源码 `internal/config/config.go`
@ -367,9 +570,11 @@ monitoring:
```
internal/
├── app/ # 应用入口、信号处理、生命周期
│ └── app.go # 主程序逻辑
│ ├── app.go # 主程序逻辑
│ └── import.go # nginx 配置导入
├── config/ # 配置加载、验证、默认值
│ ├── config.go # 配置结构定义
│ ├── loader.go # 配置文件加载
│ ├── defaults.go # 默认配置
│ └── validate.go # 配置验证
├── server/ # HTTP 服务器核心
@ -378,52 +583,114 @@ internal/
│ ├── pool.go # Goroutine 池
│ ├── status.go # 状态端点
│ ├── pprof.go # pprof 端点
│ ├── pprof_impl.go # pprof 实现
│ ├── purge.go # 缓存清除端点
│ ├── internal.go # 内部端点处理
│ └── upgrade.go # 热升级管理
├── handler/ # 请求处理器
│ ├── router.go # 路由器
│ ├── static.go # 静态文件处理
│ ├── sendfile.go # 零拷贝传输
│ ├── sendfile.go # 零拷贝传输(通用)
│ ├── sendfile_linux.go # Linux sendfile 实现
│ └── errorpage.go # 错误页面管理
├── proxy/ # 反向代理
│ ├── proxy.go # 代理核心逻辑
│ ├── websocket.go # WebSocket 代理
│ └── health.go # 健康检查
│ ├── health.go # 健康检查
│ ├── headers.go # 头部处理工具
│ ├── redirect_rewrite.go # Location 头改写
│ ├── proxy_ssl.go # SSL 代理
│ └── tempfile.go # 临时文件管理
├── loadbalance/ # 负载均衡算法
│ ├── balancer.go # 算法实现
│ └── consistent_hash.go # 一致性哈希
│ ├── balancer.go # 算法接口
│ ├── algorithms.go # 算法注册round-robin 等)
│ ├── consistent_hash.go # 一致性哈希
│ ├── random.go # Power of Two Choices
│ └── slow_start.go # 慢启动算法
├── matcher/ # Location 匹配器
│ ├── matcher.go # 匹配器接口
│ ├── exact.go # 精确匹配
│ ├── prefix.go # 前缀匹配
│ ├── regex.go # 正则匹配
│ └── radix.go # 基数树匹配
├── middleware/ # 中间件链
│ ├── middleware.go # 中间件接口
│ ├── compression/ # Gzip/Brotli 压缩
│ ├── security/ # 访问控制、限流、认证、安全头部
│ │ ├── compression.go # 压缩实现
│ │ └── gzip_static.go # 预压缩文件
│ ├── security/ # 安全中间件
│ │ ├── access.go # 访问控制
│ │ ├── ratelimit.go # 限流算法
│ │ ├── sliding_window.go # 滑动窗口
│ │ ├── auth.go # Basic Auth
│ │ ├── auth_request.go # 子请求认证
│ │ ├── geoip.go # GeoIP 过滤
│ │ └ headers.go # 安全头部
│ ├── rewrite/ # URL 重写
│ │ └ rewrite.go # 重写实现
│ ├── accesslog/ # 访问日志
│ │ └ accesslog.go # 日志实现
│ ├── bodylimit/ # 请求体大小限制
│ │ └ bodylimit.go # 限制实现
│ ├── limitrate/ # 响应速率限制
│ │ └ limitrate.go # 限速实现
│ └── errorintercept/ # 错误页面拦截
│ └ errorintercept.go # 拦截实现
├── lua/ # Lua 脚本引擎
│ ├── engine.go # gopher-lua 引擎
│ ├── context.go # Lua 请求上下文
│ └── api_*.go # nginx-lua 兼容 API
│ ├── register.go # API 注册
│ ├── socket_manager.go # Socket 管理
│ ├── filter_writer.go # header/body_filter
│ └── api_*.go # nginx-lua 兼容 API20+ 文件)
├── http2/ # HTTP/2 服务器
│ ├── server.go # HTTP/2 服务器实现
│ └── adapter.go # HTTP/2 适配器
├── http3/ # HTTP/3 服务器
│ ├── server.go # QUIC 服务器
│ └── adapter.go # HTTP/3 适配器
├── stream/ # TCP/UDP Stream 代理
│ └── stream.go # 四层代理实现
│ ├── stream.go # 四层代理实现
│ └ ssl.go # Stream SSL 支持
├── ssl/ # TLS 配置
│ ├── ssl.go # TLS 管理器
│ ├── ocsp.go # OCSP Stapling
│ └── session_tickets.go # Session Tickets 密钥轮换
│ ├── session_tickets.go # Session Tickets 密钥轮换
│ └ client_verify.go # 客户端证书验证
├── cache/ # 缓存系统
│ ├── backend.go # 缓存后端抽象
│ ├── file_cache.go # 文件缓存
│ └── purge.go # 代理缓存
│ ├── disk_cache.go # 磁盘缓存
│ ├── tiered_cache.go # L1/L2 分层缓存
│ └ purge.go # 缓存清除
├── resolver/ # DNS 解析器
│ └── resolver.go # TTL 缓存、异步解析
│ ├── resolver.go # DNS 解析逻辑
│ ├── cache.go # DNS 结果缓存
│ └ stats.go # DNS 解析统计
├── variable/ # 变量系统
│ └── variable.go # 日志/头部变量
│ ├── variable.go # 变量接口
│ ├── builtin.go # 内置变量
│ ├── pool.go # 变量池
│ └ ssl.go # SSL 相关变量
├── converter/ # 配置转换器
│ └ nginx/ # nginx 配置导入
│ ├── parser.go # nginx 语法解析
│ └ converter.go # YAML 转换
├── logging/ # 日志系统
│ └── logging.go # 结构化日志zerolog
│ └ logging.go # 结构化日志zerolog
├── netutil/ # 网络工具
│ ├── ip.go # IP 解析
│ └── url.go # URL 处理
│ ├── host.go # 主机名处理
│ └ url.go # URL 处理
├── mimeutil/ # MIME 类型检测
│ └ detect.go # MIME 检测实现
├── sslutil/ # SSL 工具
│ └ certpool.go # 证书池管理
├── utils/ # 通用工具
│ ├── httperror.go # HTTP 错误处理
│ └ internal.go # 内部工具
├── version/ # 版本信息
│ └ version.go # 版本定义
└── benchmark/ # 基准测试工具
└── tools/ # 测试辅助工具
```
@ -495,28 +762,103 @@ ngx.log(ngx.ERR, "error message")
ngx.log(ngx.WARN, "warning message")
ngx.log(ngx.INFO, "info message")
-- 请求操作
-- HTTP 状态码常量
ngx.HTTP_OK = 200
ngx.HTTP_NOT_FOUND = 404
ngx.HTTP_INTERNAL_SERVER_ERROR = 500
-- 更多: HTTP_MOVED_PERMANENTLY, HTTP_BAD_REQUEST, HTTP_FORBIDDEN 等
-- 特殊常量
ngx.OK, ngx.ERROR, ngx.AGAIN, ngx.DONE, ngx.DECLINED
-- 内置变量ngx.var
local uri = ngx.var.uri
local args = ngx.req.get_uri_args()
ngx.req.set_uri("/new_path")
local method = ngx.var.request_method
local remote_addr = ngx.var.remote_addr
local host = ngx.var.host
local request_uri = ngx.var.request_uri
local scheme = ngx.var.scheme
local server_port = ngx.var.server_port
-- 任意请求头: ngx.var.http_xxx如 ngx.var.http_user_agent
-- 任意查询参数: ngx.var.arg_xxx如 ngx.var.arg_token
-- 请求操作ngx.req
ngx.req.get_method() -- 获取 HTTP 方法
ngx.req.get_uri() -- 获取 URI 路径
ngx.req.set_uri("/new_path") -- 设置 URI
ngx.req.set_uri_args("a=1&b=2") -- 设置查询参数
ngx.req.get_uri_args() -- 获取查询参数表
ngx.req.get_headers(max?) -- 获取请求头表
ngx.req.set_header("X-Custom", "value")
ngx.req.clear_header("X-Custom")
ngx.req.get_body_data() -- 获取请求体
ngx.req.read_body() -- 确保请求体已读取
-- 响应操作
ngx.header["X-Response"] = "value"
ngx.say("Hello World")
ngx.exit(200)
-- 响应操作ngx.resp
ngx.resp.get_status() -- 获取响应状态码
ngx.resp.set_status(200) -- 设置响应状态码
ngx.resp.get_headers(max?) -- 获取响应头表
ngx.resp.set_header("X-Response", "value")
ngx.resp.clear_header("X-Response")
-- 共享字典
-- 快捷响应
ngx.say("Hello World") -- 输出内容并换行
ngx.print("Hello World") -- 输出内容不换行
ngx.flush(wait?) -- 刷新响应缓冲区
ngx.redirect("/new", 302) -- 重定向
ngx.exit(200) -- 结束请求处理
-- 请求上下文ngx.ctx每请求独立的 table
ngx.ctx.user_id = 123
local id = ngx.ctx.user_id
-- 共享字典ngx.shared.DICT
local dict = ngx.shared.DICT
dict:set("key", "value", 3600)
local val = dict:get("key")
dict:set("key", "value", 3600) -- 设置(含过期时间)
dict:add("key", "value", 3600) -- 仅键不存在时添加
dict:replace("key", "value") -- 仅键存在时替换
local val = dict:get("key") -- 获取值
dict:incr("counter", 1) -- 数值自增
dict:delete("key") -- 删除键
dict:flush_all() -- 清空所有条目
dict:flush_expired(max?) -- 清除过期条目
dict:get_keys(max?) -- 获取所有键
dict:size() -- 获取条目数
dict:free_space() -- 获取剩余容量
-- 元表语法: dict["key"] = value / val = dict["key"]
-- Socket受限
-- Socketngx.socket.tcp受限)
local sock = ngx.socket.tcp()
sock:connect("127.0.0.1", 8080)
sock:settimeout(5000) -- 设置超时(毫秒)
sock:settimeouts(1000, 2000, 3000) -- 分别设置连接/发送/读取超时
sock:send("GET / HTTP/1.0\r\n\r\n")
local response = sock:receive("*a")
local response = sock:receive("*a") -- 读取全部
local line = sock:receive("*l") -- 读取一行
local reader = sock:receiveuntil("--boundary") -- 按模式读取
sock:close()
-- 定时器ngx.timer
local timer_id = ngx.timer.at(5, function()
ngx.log(ngx.INFO, "timer fired")
end)
ngx.timer.running_count() -- 获取运行中的定时器数量
-- handle:cancel() -- 取消定时器(通过返回的 handle
-- 子请求ngx.location.capture
local res = ngx.location.capture("/internal/auth", {
method = ngx.HTTP_GET,
body = "request body",
args = "a=1&b=2"
})
-- res.status, res.body, res.headers
-- 动态负载均衡ngx.balancer
ngx.balancer.set_current_peer("backend:8080") -- 设置后端地址
ngx.balancer.set_more_tries(3) -- 设置重试次数
local fail_type = ngx.balancer.get_last_failure() -- 获取上次失败类型
local targets = ngx.balancer.get_targets() -- 获取所有可用目标
local client_ip = ngx.balancer.get_client_ip() -- 获取客户端 IP
```
### 示例脚本
@ -668,44 +1010,57 @@ sudo systemctl status lolly # 查看状态
```json
{
"connections": {
"active": 150,
"reading": 2,
"writing": 10,
"idle": 138
"version": "0.2.1",
"uptime": "2h30m15s",
"connections": 150,
"requests": 125000,
"bytes_sent": 5368709120,
"bytes_received": 1073741824,
"pool": {
"workers": 256,
"IdleWorkers": 200,
"MaxWorkers": 10000,
"MinWorkers": 100,
"QueueLen": 0,
"QueueCap": 1000
},
"requests": {
"total": 125000,
"current": 150,
"per_second": 1250
},
"traffic": {
"in_bytes": 1073741824,
"out_bytes": 5368709120
},
"upstreams": {
"backend1": {
"active": 50,
"healthy": true,
"total_requests": 80000
},
"backend2": {
"active": 30,
"healthy": true,
"total_requests": 45000
"upstreams": [
{
"name": "backend1",
"targets": [
{"url": "http://10.0.0.1:8080", "healthy": true, "latency_ms": 5},
{"url": "http://10.0.0.2:8080", "healthy": true, "latency_ms": 8}
],
"healthy_count": 2,
"unhealthy_count": 0,
"latency_p50_ms": 5.2,
"latency_p95_ms": 12.5,
"latency_p99_ms": 25.0
}
],
"rate_limits": [
{
"zone_name": "api_limit",
"requests": 50000,
"limit": 100,
"rejected": 150
}
],
"ssl": {
"handshakes": 2500,
"session_reused": 800,
"reuse_rate_percent": 32.0
},
"cache": {
"file_cache": {
"entries": 5000,
"size_bytes": 134217728,
"hits": 85000,
"misses": 15000
"max_entries": 50000,
"size": 134217728,
"max_size": 268435456
},
"proxy_cache": {
"entries": 2000,
"hits": 70000,
"misses": 10000
"pending": 5
}
}
}
@ -961,7 +1316,7 @@ curl http://localhost:8080/status | jq '.cache'
### 环境要求
- Go 1.26+
- Go 1.24+
- make
### 命令
@ -994,7 +1349,7 @@ make lint
### 项目统计
- Go 文件:132
- Go 文件290
- 测试文件157
- 核心模块均有完整测试和性能基准测试
- 中文代码注释
@ -1004,11 +1359,20 @@ make lint
| 库 | 版本 | 用途 |
|------|---------|------|
| [fasthttp](https://github.com/valyala/fasthttp) | v1.70.0 | HTTP 服务器核心 |
| [fasthttp/router](https://github.com/fasthttp/router) | v1.5.4 | HTTP 路由器 |
| [quic-go](https://github.com/quic-go/quic-go) | v0.59.0 | QUIC/HTTP/3 实现 |
| [zerolog](https://github.com/rs/zerolog) | v1.35.0 | 零分配 JSON 日志 |
| [gopher-lua](https://github.com/yuin/gopher-lua) | v1.1.2 | Lua 脚本引擎 |
| [klauspost/compress](https://github.com/klauspost/compress) | v1.18.5 | Gzip/Brotli 压缩 |
| [brotli](https://github.com/andybalholm/brotli) | v1.2.1 | Brotli 压缩编码 |
| [geoip2-golang](https://github.com/oschwald/geoip2-golang) | v1.13.0 | GeoIP 查询 |
| [golang-lru/v2](https://github.com/hashicorp/golang-lru) | v2.0.7 | LRU 缓存 |
| [uuid](https://github.com/google/uuid) | v1.6.0 | UUID 生成 |
| [testcontainers-go](https://github.com/testcontainers/testcontainers-go) | v0.42.0 | 集成测试容器 |
| [testify](https://github.com/stretchr/testify) | v1.11.1 | 测试断言 |
| [golang.org/x/crypto](https://golang.org/x/crypto) | v0.50.0 | 加密工具 |
| [golang.org/x/net](https://golang.org/x/net) | v0.53.0 | 网络扩展 |
| [yaml.v3](https://gopkg.in/yaml.v3) | v3.0.1 | YAML 配置解析 |
## 许可证