f4nff

``` gost -L=ws://:80?out=500k&in=500k gost -L=:1080 -F=ws://:80?out=500k&in=500k ``` 针对单连接进行最大带宽限制，使用iperf3进行测速，如果限制不超过某个值，传输非常平稳， 如果超过某个值之后，会被Qos惩戒，然后当前连接速率就为0，只能断开，重新发起连接，才有速率。 客户端限制或者服务端限制，同时有效。

websocket添加ping pongs，实现双向心跳， http2同理， 不然在nat层，没有数据传输，常常会被断开连接， 还有添加限流，针对单连接限制最大上行，下行速率，

``` root@yy:/home/gopkg/src/github.com/FlowerWrong/tun2socks/cmd# go get -d github.com/FlowerWrong/tun2socks/cmd root@yy:/home/gopkg/src/github.com/FlowerWrong/tun2socks/cmd# ls main.go netstack shared smart surge udp root@yy:/home/gopkg/src/github.com/FlowerWrong/tun2socks/cmd# go build # github.com/google/netstack/tcpip/link/rawfile ../../../google/netstack/tcpip/link/rawfile/rawfile_unsafe.go:137:10: undefined: BlockingPoll ../../../google/netstack/tcpip/link/rawfile/rawfile_unsafe.go:159:10: undefined: BlockingPoll ../../../google/netstack/tcpip/link/rawfile/rawfile_unsafe.go:189:14: undefined: BlockingPoll # github.com/google/netstack/tcpip/stack...

Investigate how often clients in the wild regenerate UDP connection IDs/decide what to do

24

The udp protocol's tracer server appears error code [bad connection ID]  `udp://37.1.210.233:80/announce` Serving chihaya，Unfortunately, he has made a mistake. `udp://ipv4.opentracker.xyz:80/announce` is [opentracker](http://erdgeist.org/arts/software/opentracker/) Opentracker is normal. please

openwrt-22.03.0-rc1-rockchip-armv8-friendlyarm_nanopi-r4s-ext4 Upstream speed limit（bug）

10

[https://downloads.openwrt.org/releases/22.03.0-rc1/targets/rockchip/armv8/openwrt-22.03.0-rc1-rockchip-armv8-friendlyarm_nanopi-r4s-ext4-sysupgrade.img.gz](https://downloads.openwrt.org/releases/22.03.0-rc1/targets/rockchip/armv8/openwrt-22.03.0-rc1-rockchip-armv8-friendlyarm_nanopi-r4s-ext4-sysupgrade.img.gz)    Upstream speed: 3Mbps The same network, I use x86 img upload speed is no problem. Only r4s openwrt has such a problem. [https://download.friendlyelec.com/nanopir4s](https://download.friendlyelec.com/nanopir4s) This is the...

debug

1

``` Ping http://127.0.0.1:9999(127.0.0.1:9999) - http is open - time=364.1457ms method=GET status=403 bytes=16 runtime: out of memory: cannot allocate 32768-byte block (1781923840 in use) fatal error: out of memory runtime: out...

ipv6 no~

2

delete

bug~！

2

带宽峰值超过某个值，大概200Mbps之后，客户端卡死了，连不上了， 但是与服务端的链接状态还存在 但是不通了， 必须重启客户端， 重现方式，找个网络质量好的，然后设置几个隧道， 然后找个youtube 4k视频，随意多次拖动就可以很快触发，

not enough arguments in call to certmagic.NewDefault().RevokeCert

1

``` root@ubuntu:/home/gopkg/src/github.com/pieterlouw/caddy-net/caddynet# go build caddynet.go # github.com/caddyserver/caddy/caddytls /home/gopkg/pkg/mod/github.com/caddyserver/[email protected]/caddytls/tls.go:80:42: not enough arguments in call to certmagic.NewDefault().RevokeCert have (string, bool) want (context.Context, string, bool) root@ubuntu:/home/gopkg/src/github.com/pieterlouw/caddy-net/caddynet# ```

负载均衡的一点想法，

2

目前负载均衡有三种模式，轮询，随机，递进。 配置如下 ``` # strategy for node selecting strategy round max_fails 2 fail_timeout 30s # period for live reloading reload 10s # peers peer mwss://ra.elyzc.net:443?ip=80.63.188.5 peer mwss://ra.qdlyzc.net:443?ip=80.63.188.6 ``` 我看了下nginx里面的负载均衡的算法貌似更合理。 ```...