TP99 latency represents the longest time required to complete 99% of network requests. Assuming that the total number of requests is 100, their request time is 0.01s, 0.02s, 0.03s, 0.04s...0.98s. .0. 99s, 1s, so TP99 is 0.99s.

Because edge nodes often have resource constraints, the IEF service provides a lightweight runtime framework to meet the requirements of fast startup, timely response, and small resource usage at the edge. Application support functions, containers, and virtual machines are three types of Runtime forms.

Huawei cloud storage disaster recovery service can meet the user's requirement to achieve RPO=0.

IEF (Intelligent EdgeFabrie) intelligent edge platform can sink the core cloud services of HUAWEI CLOUD such as (AI services, big data services, integration services, etc.) to edge nodes. Regular inspection of services is an extension of HUAWEI CLOUD services at the edge, which complements HUAWEI CLOUD.

Models trained by ModelArts can only be deployed on online services based on ModelArts.