-
-
Notifications
You must be signed in to change notification settings - Fork 14.7k
Open
Description
Bug Description
之前服务端是0.5x.x,客户端是0.49.0,用了很久,没有太大问题,但是遇到长连接数超过一定数量的就会出现404导致服务崩溃,必须手动重启frps才能解决。
然后我更新到0.65.0,现在非常频繁的出现404,哪怕极低负载也会频繁出现404,一会好一会404,间歇性的。
两个版本更新前后用的配置还是有区别的,之前用的ini,后面改成toml,配置项我已经不记得之前是怎么写的了。
以下是新的配置。
frpc Version
0.65.0
frps Version
0.65.0
System Architecture
linux/amd64
Configurations
frps:
# frps.toml - 正确配置
bindAddr = "0.0.0.0"
bindPort = 5443
kcpBindPort = 5443
vhostHTTPPort = 7860
# 核心调整:将此参数设为12分钟(720秒),解决HTTP请求排队问题
vhostHTTPTimeout = 720
# 连接保活设置:恢复合理值,确保连接通道的健康
transport.heartbeatTimeout = 90 # 恢复为默认的90秒或稍大值(如120秒)
transport.tcpKeepalive = 7200 # 保持默认的2小时,不要修改
# 连接池与多路复用设置(您的现有配置很好,继续保持)
transport.maxPoolCount = 100
transport.tcpMux = true
transport.tcpMuxKeepaliveInterval = 60 # 60秒是很好的设置
auth.method = "token"
auth.token = "8ad3d13x429..."
webServer.addr = "0.0.0.0"
webServer.port = 9527
webServer.user = "admin"
webServer.password = "test..."
allowPorts = [
{ start = 1024, end = 65535 }
]
log.to = "console"
log.level = "info"
frpc:
# ========================
# 基础服务端连接配置
# ========================
serverAddr = "hk.77chat.cn"
serverPort = 5443
# 身份认证 (必须与服务端一致)
auth.method = "token"
auth.token = "8adfdfd..."
# ========================
# 客户端全局优化配置
# ========================
# 1. 防崩溃机制:登录失败不退出,持续重试
loginFailExit = false
# 2. 传输层优化
[transport]
heartbeatTimeout = 90 # 与服务端同步,维持连接活性
tcpMux = true # 启用多路复用,提升性能
tls.enable = false
# 3. 日志配置
#[log]
#to = "./frpc.log"
#level = "info"
#maxDays = 3
# ========================
# 代理配置 (Services/Proxies)
# ========================
Logs
No response
Steps to reproduce
...
Affected area
- Docs
- Installation
- Performance and Scalability
- Security
- User Experience
- Test and Release
- Developer Infrastructure
- Client Plugin
- Server Plugin
- Extensions
- Others
Metadata
Metadata
Assignees
Labels
No labels