Noticed while investigating https://issues.redhat.com/browse/OSPCIX-517 that even after bumping haproxy timeout, some of the api request timed out. And this due to the services running with httpd do not have timeout configurable(default is 60 seconds)

Some of the service operators already have it configurable, like:-

manila, cinder, glance

https://github.com/openstack-k8s-operators/glance-operator/pull/550

https://github.com/openstack-k8s-operators/cinder-operator/pull/396

https://github.com/openstack-k8s-operators/manila-operator/pull/282

And followup on openstack-operator https://github.com/openstack-k8s-operators/openstack-operator/pull/830

Some have hard coded but higher timeouts
barbican 90 secondsheat 600 seconds
And others don't have it configurable:-
nova
neutron
telemetry
swift
ironic
octavia
keystone
horizon
placement
designate
 
This ticket is to track the work for the service which still not have it configurable, will create bug ticket for each of the operator