# Http health check with SSL backend

**URL:** <https://discourse.haproxy.org/t/http-health-check-with-ssl-backend/703>\
**Category:** Help!\
**Created:** [October 5, 2016, 9:05pm UTC](https://discourse.haproxy.org/t/http-health-check-with-ssl-backend/703 "2016-10-05T21:05:41Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![cavemandaveman](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.haproxy.org/cavemandaveman/32/123_2.png) [@cavemandaveman](https://discourse.haproxy.org/u/cavemandaveman)\
**Post date:** [October 5, 2016, 9:05pm UTC](https://discourse.haproxy.org/t/http-health-check-with-ssl-backend/703/1 "2016-10-05T21:05:41Z")

</div>

Hello, I’m having a problem with haproxy health checks.

I have a backend that is serving the main service on HTTPS 8443 and serving a health endpoint on HTTP 9000.

Is there a way to have the haproxy health check over http while the backend server is https? My current config (which does not work) looks like this:

```
backend my_backend
  mode http
  timeout check 2000
  option httpchk GET "/health" "HTTP/1.0"
  cookie my-cookie insert nocache postonly domain example.com
  server my_server 10.42.38.175:8443 ssl verify none check port 9000 inter 2000 rise 2 fall 3 cookie my_server
  http-request add-header X-Forwarded-Proto https if { ssl_fc }
  http-request set-header X-Forwarded-Port %[dst_port]

```

I’m thinking the issue is that because there is the `ssl` option on the server line, it’s expecting everything to be SSL, even the `check port`? Is there a way around this?

---

<div class="post-metadata">

**Author:** ![cavemandaveman](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.haproxy.org/cavemandaveman/32/123_2.png) [@cavemandaveman](https://discourse.haproxy.org/u/cavemandaveman)\
**Post date:** [October 6, 2016, 6:17pm UTC](https://discourse.haproxy.org/t/http-health-check-with-ssl-backend/703/2 "2016-10-06T18:17:19Z")

</div>

Please ignore this thread. The above config _does_ work.

The root cause of my issue was due to the fact that haproxy closes the health check connection with RST instead of FIN, and therefore I was seeing `ClientAbortException` errors in my backend application. Totally different topic.