开启负载均衡SLB后ECS存在大量TIMEWAIT的内网80端口连接

负载均衡默认会抹除 tcp 连接的 timestamp 属性，linux协议栈的 tw_reuse(time_wait 状态连接复用)无法生效，time_wait 状态连接堆积导致客户端端口不足

解决方法：

客户端端使用长连接代替短连接。

使用 RST 报文断开连接（socket 设置 SO_LINGER 属性） ，而不是发 FIN 包这种方式断开。

解决了嘛

这个问题有解决吗， 我们也注意到了大量sLB健康检查导致的TIME_WAIT连接

不要这样设置：
net.ipv4.tcp_timestamps = 1
net.ipv4.tcp_tw_reuse = 1
net.ipv4.tcp_tw_recycle = 1
这样会导致局域网访问的用户可能失败。

给一份我的线上配置：
fs.nr_open = 5242880
fs.file-max = 4194304
kernel.core_uses_pid = 1
kernel.msgmax = 1048560
kernel.msgmnb = 1073741824
kernel.shmall = 4294967296
kernel.shmmax = 68719476736
kernel.sysrq = 0
net.core.netdev_max_backlog = 1048576
net.core.rmem_default = 2097152
net.core.rmem_max = 16777216
net.core.somaxconn = 1048576
net.core.wmem_default = 2097152
net.core.wmem_max = 16777216
net.ipv4.conf.default.accept_source_route = 0
net.ipv4.conf.default.rp_filter = 1
net.ipv4.ip_forward = 1
net.ipv4.ip_local_port_range = 1024    65533
net.ipv4.neigh.default.gc_thresh1 = 10240
net.ipv4.neigh.default.gc_thresh2 = 40960
net.ipv4.neigh.default.gc_thresh3 = 81920
net.nf_conntrack_max = 819200
net.netfilter.nf_conntrack_max = 819200
net.netfilter.nf_conntrack_tcp_timeout_close_wait = 60
net.netfilter.nf_conntrack_tcp_timeout_fin_wait = 120
net.netfilter.nf_conntrack_tcp_timeout_time_wait = 120
net.ipv4.tcp_fin_timeout = 10
net.ipv4.tcp_keepalive_intvl = 15
net.ipv4.tcp_keepalive_probes = 5
net.ipv4.tcp_keepalive_time = 30
net.ipv4.tcp_max_orphans = 3276800
net.ipv4.tcp_max_syn_backlog = 1048576
net.ipv4.tcp_max_tw_buckets = 60000
net.ipv4.tcp_mem = 94500000 915000000 927000000
net.ipv4.tcp_orphan_retries = 3
net.ipv4.tcp_reordering = 5
net.ipv4.tcp_retrans_collapse = 0
net.ipv4.tcp_retries2 = 5
net.ipv4.tcp_rmem = 4096        87380   4194304
net.ipv4.tcp_sack = 0
net.ipv4.tcp_synack_retries = 1
net.ipv4.tcp_syncookies = 0
net.ipv4.tcp_syn_retries = 1
net.ipv4.tcp_timestamps = 0
net.ipv4.tcp_tw_recycle = 0
net.ipv4.tcp_tw_reuse = 1
net.ipv4.tcp_window_scaling = 1
net.ipv4.tcp_wmem = 4096        16384   4194304
net.ipv6.conf.all.disable_ipv6 = 1
net.ipv6.conf.default.disable_ipv6 = 1
vm.swappiness = 0
net.ipv4.tcp_low_latency = 1

我也遇到这个问题了，不知道如何解决

开启负载均衡SLB后ECS会存在大量TIMEWAIT的内网80端口连接(几秒内升到2W预设值)，不知道如何解决，不走SLB通过另一个ECS压力测试时，仅会出现极少量(低于500)的TIMEWAIT

ECS的sysctl.conf配置如下：

net.ipv4.ip_forward = 0
net.ipv4.conf.all.rp_filter = 0
net.ipv4.conf.default.rp_filter = 0
net.ipv4.conf.default.accept_source_route = 0
kernel.sysrq = 0
kernel.core_uses_pid = 1
kernel.msgmnb = 65536
kernel.msgmax = 65536
kernel.shmmax = 68719476736
kernel.shmall = 4294967296
net.ipv4.neigh.default.gc_stale_time = 120
net.ipv4.conf.default.arp_announce = 2
net.ipv4.conf.all.arp_announce = 2
net.ipv4.conf.lo.arp_announce = 2

net.ipv4.tcp_fin_timeout = 30
net.ipv4.tcp_keepalive_time = 300
net.ipv4.tcp_syncookies = 1

net.ipv4.tcp_timestamps = 1
net.ipv4.tcp_tw_reuse = 1
net.ipv4.tcp_tw_recycle = 1

net.ipv4.ip_local_port_range = 1024 65000
net.ipv4.tcp_max_tw_buckets = 20000
net.core.somaxconn = 81920
net.core.netdev_max_backlog = 81920
net.ipv4.tcp_max_syn_backlog = 81920
net.ipv4.tcp_synack_retries = 3
net.ipv4.tcp_max_orphans = 3276800
net.ipv4.tcp_no_metrics_save = 1
net.ipv4.tcp_sack = 1
net.ipv4.tcp_window_scaling = 1

压力测试命令为：

ab -k -c10000 -n400000 url

-------------------------

目前SLB服务器的机器访问ECS全部采用的是http 1.0短链接协议，而真实用户访问SLB时采用的是http 1.1，经过SLB服务器中转之后变为了http 1.0，我猜这是导致大量TIME_WAIT的一个主要原因！

果然

是不是健康检查啊？