As noted on an earlier patch[1], the "SetAmphoraFirewallRules" task was not
checking the Amphora status nor using an API timeout. This could cause failover
flows to take longer than necessary if one of the Amphora is missing.
This patch corrects that issue by honoring both the Amphora status and timeout.
[1] https://review.opendev.org/c/openstack/octavia/+/910101/13/octavia/controller/worker/v2/flows/amphora_flows.py
Change-Id: Ic5e8140b13164267236f0a5d9a48fbd84bcdd688