craned cpu utilization continues at 100% and resource and replicas recommended values are older

[Cloudzp]

Describe the bug
When using craned for Resource and Replicas recommendation in a larger cluster, when the number of workloads reaches 2w+, the CPU utilization continues to be 100% (limit 8c), resource and replicas recommended values are older( runInterval: 24h0m0s)
Reproduce steps

1. Create more than 20,000 workload objects;
2. Configure recommended rules for all workloads;

Expected behavior
The Resource and Replicas recommendation work properly

Screenshots

Environment (please complete the following information):

• K8S Version: [e.g. v1.14.3]
• Crane Version: [e.g. pr-872-bb5c1e4]