* Previously, if you configured an {product-title} cluster with a high number of internal services or user-managed load balancer IP addresses, you experienced a delayed startup time for the OVN-Kubernetes service. This delay occurred when the OVN-Kubernetes service attempted to install `iptables` rules on a node. With this release, the OVN-Kubernetes service can process a large number of services in a few seconds. Additionally, you can access a new log to view the status of installing `iptables` rules on a node. (link: https://issues.redhat.com/browse/OCPBUGS-34273 [* OCPBUGS-34273 *]) _________ *Cause*: What actions or circumstances cause this bug to present. When ovnkube-node pod is started and there are many external IPs or load balancer IP services exist. *Consequence*: What happens when the bug presents. ovnkube-node will take a very long time to start up due to it programming iptables rules one at a time for all services. Iterating over external IPs or load balancer iPs is especially costly. The length of start up time depends on the number of services and external IPs/load balancer IPs. We have seen this time take from several minutes to almost an hour. *Fix*: What was done to fix the bug. Optimizations in the service parsing logic reduces duplicate calls to create iptables rules. Now iptables rules are atomically created using iptable-restore rather than a separate call to iptables for each external IP. *Result*: Bug doesn’t present anymore. The start up time is much faster, even at high scale it should be on the order of seconds rather than minutes.