A financial company must log and retain all internet service traffic from any host for two years. In development they use VPC Traffic Mirroring with a Network Load Balancer (NLB) as the mirror target and mirror traffic into another account. When deployed in production, not all traffic is mirrored and the loss appears random. Which explanations could cause some traffic not to be mirrored? (Choose two.)

Mirrored traffic is assigned lower priority by the underlying network infrastructure. During periods of network congestion, the system prioritizes production traffic, leading to mirrored traffic being dropped. This explains why the loss appears random and not all traffic is mirrored. The NLB experiencing warm-up latency due to a sudden and significant traffic increase can also cause intermittent traffic loss. NLBs need time to scale up to handle unexpected spikes in traffic, and during this warm-up period, some connections might be dropped or not processed, including mirrored traffic. Security group misconfigurations would likely block all traffic, not just some mirrored traffic, and would typically manifest as a complete failure rather than random loss. An IAM policy misconfiguration would prevent the creation of mirror sessions entirely, not cause partial mirroring. While EC2 instances can be overloaded, Traffic Mirroring is performed at the ENI level and doesn't directly add load to the monitored instance's CPU or memory, so it wouldn't cause the instance to fail to keep up.