W dniu 2017-12-29 o 15:28, Lukas Tribus pisze:
> Hello,
>
>
> On Fri, Dec 29, 2017 at 3:24 PM, Mariusz Kalota <[email protected]> wrote:
>> Hello, thanks for reply.
>>
>> I would like to forward the encrypted HTTPS transparently to the
>> backend. I would like to check health my backend servers, but not only
>> on layer 4, but on layer 7. So I have to get /somefile.asmx, and I have
>> to give client certificate, and I have to check the backend server not
>> using IP, but using hostname.
>>
>> The two backend servers are IIS server, which have binding's:
>>
>> server1:
>> test.site.local:51111
>> test1.site.local:51111
>>
>> server2:
>> test.site.local:51111
>> test2.site.local:51111
>>
>> If I use in webbrowser test.site.local I go to haproxy. If I use
>> test1.site.local or test2.site.local I go to one of the backend server
>> directly.
>>
>> So currently my main issue is to check healhy of backend servers on
>> layer 7 (http 200 response).
> Understood, please follow the suggestions from the previous mail. Also
> make sure sni and host header are set to the same string.
>
>
> Lukas
Hello,
there is some progres. My current config:

listen https-test-51111
bind *:51111
mode tcp
balance roundrobin
option httpchk GET /somefile.asmx test2.site.local:51111
server server1 192.168.0.1:51111 check check-ssl verify none crt 
/etc/haproxy/cert.pem ssl sni str(test1.site.local) check-sni test1.site.local
server server2 192.168.0.2:51111 check check-ssl verify none crt 
/etc/haproxy/cert.pem ssl sni str(test2.site.local) check-sni test2.site.local


haproxy stats:
server1 L7STS/400 in 3ms
server2 L7OK/200 in 26ms

The second server is ok, because there is

option httpchk GET /somefile.asmx test2.site.local:51111

but the first not.

How can I fix this?

Regards,
Mariusz



Reply via email to