# Haproxy 1.8.3 seamless reload refresh health checks

**URL:** <https://discourse.haproxy.org/t/haproxy-1-8-3-seamless-reload-refresh-health-checks/2108>\
**Category:** Help!\
**Created:** [February 9, 2018, 4:24pm UTC](https://discourse.haproxy.org/t/haproxy-1-8-3-seamless-reload-refresh-health-checks/2108 "2018-02-09T16:24:29Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![cuq70633](https://avatars.discourse-cdn.com/v4/letter/c/f1d935/32.png) [@cuq70633](https://discourse.haproxy.org/u/cuq70633)\
**Post date:** [February 9, 2018, 4:24pm UTC](https://discourse.haproxy.org/t/haproxy-1-8-3-seamless-reload-refresh-health-checks/2108/1 "2018-02-09T16:24:29Z")

</div>

Hello All.  
When haproxy seamless reload config file, all health checks refresh to state “ok”, so backends become UP and receiving connections for few seconds.

Haproxy starts by command:  
haproxy -D -f /usr/local/etc/haproxy/haproxy.cfg -p /var/run/haproxy.pid -x /var/run/haproxy.sock -sf $(cat /var/run/haproxy.pid)

Haproxy cfg for nbproc \>1 has string:  
stats socket /var/run/haproxy.sock mode 660 level admin expose-fd listeners process 1

and health check strings:

---

<div class="post-metadata">

**Author:** ![happy](https://avatars.discourse-cdn.com/v4/letter/h/f05b48/32.png) [@happy](https://discourse.haproxy.org/u/happy)\
**Post date:** [September 22, 2018, 4:30am UTC](https://discourse.haproxy.org/t/haproxy-1-8-3-seamless-reload-refresh-health-checks/2108/2 "2018-09-22T04:30:02Z")

</div>

1. Before reload:  
dump the current states to file, using haproxy stats socket and command “show servers state”
2. Set haproxy.cfg to read this file like:  
server-state-file-name /tmp/haproxy\_server\_states  
load-server-state-from-file local