这是我的haproxy.conf(haproxy 1.7.9)
global
log 127.0.0.1 local0
defaults
retries 3
option redispatch
timeout client 30s
timeout connect 30s
timeout server 30s
option http-keep-alive
http-reuse always
frontend web1
bind *:8080
option http-keep-alive
mode http
default_backend app1
backend app1
balance roundrobin
option http-keep-alive
mode http
server a2 192.168.56.150:8000
原始服务器的卷曲结果:
$ curl -vv http://192.168.56.150:8000/test --keepalive-time 700
* About to connect() to 192.168.56.150 port 8000 (#0)
* Trying 192.168.56.150...
* Connected to 192.168.56.150 (192.168.56.150) port 8000 (#0)
> GET /test HTTP/1.1
> User-Agent: curl/7.29.0
> Host: 192.168.56.150:8000
> Accept: */*
>
< HTTP/1.1 200 OK
< Etag: 720-6-59eeda80
< Content-Type: application/octet-stream
< Content-Length: 6
< Last-Modified: Tue, 24 Oct 2017 06:15:28 GMT
< Server: WEBrick/1.3.1 (Ruby/2.0.0/2015-12-16)
< Date: Fri, 27 Oct 2017 02:38:14 GMT
< Connection: Keep-Alive
<
tests
* Connection #0 to host 192.168.56.150 left intact
haproxy服务器的卷曲结果
$ curl -vv http://192.168.56.150:8080/test --keepalive-time 700
* About to connect() to 192.168.56.150 port 8080 (#0)
* Trying 192.168.56.150...
* Connected to 192.168.56.150 (192.168.56.150) port 8080 (#0)
> GET /test HTTP/1.1
> User-Agent: curl/7.29.0
> Host: 192.168.56.150:8080
> Accept: */*
>
< HTTP/1.1 200 OK
< Etag: 720-6-59eeda80
< Content-Type: application/octet-stream
< Content-Length: 6
< Last-Modified: Tue, 24 Oct 2017 06:15:28 GMT
< Server: WEBrick/1.3.1 (Ruby/2.0.0/2015-12-16)
< Date: Fri, 27 Oct 2017 02:38:05 GMT
<
tests
* Connection #0 to host 192.168.56.150 left intact
我在tcpdump
确认从原点到haproxy的回复中有Connection: Keep-Alive
正如你所看到的,haproxy响应中没有Connection: Keep-Alive
,我怎样才能使haproxy保持活着?
答案 0 :(得分:1)
默认情况下,HTTP 1.1以保持活动模式运行,请参阅RFC7230。因此,您不需要明确设置连接标头以使用持久连接(保持活动模式)。
仅在使用HTTP 1.0时才需要连接头,因为它旨在在每次请求后关闭连接。
因此,要验证HAProxy是否处于保持活动模式,您需要使用curl发送多个HTTP请求(而不仅仅是一个),并检查是否只建立并使用了一个连接。
这个serverfault post显示了如何做到这一点。 (注意&#34;正在连接...&#34; 和&#34;关闭连接&#34; 行)