Using mod_proxy_balancer some client requests receive 503 response
Issue
- Testing a disaster recovery by making a node unavailable in a
mod_proxy_balancerconfiguration similar to the following results in some clients receiving 503 responses:
<Proxy balancer://cluster1>
BalancerMember http://member1:8180 route=member1 max=20 ttl=120
BalancerMember http://member2:8180 route=member2 max=20 ttl=120
ProxySet lbmethod=byrequests
ProxySet stickysession=JSESSIONID
ProxySet nofailover=Off
</Proxy>
- Clients are not stuck with a down node, however some requests are made and bring a 503 to clients
Environment
- Red Hat Enterprise Linux
- 7.x
- 8.x
mod_proxy_balancer
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.