KubeSphere版本：v3.4.1

目前通过api取内存使用率，感觉数值有点不对，麻烦大佬解答下

http://1.1.1.1:30100/kapis/monitoring.kubesphere.io/v1alpha3/namespaces/myns/pods/mypod-69495fbd5-jwjnp

metrics_filter=pod_memory_usage

最后取出来两个内存使用，如下

“metric_name”: “pod_memory_usage”

  "values": [

   [

    1750694400,

    "4529881088"

   ],

   [

    1750698000,

    "4531036160"

   ],

“metric_name”: “pod_memory_usage_wo_cache”,

  "values": [

   [

    1750694400,

    "1898369024"

   ],

   [

    1750698000,

    "1894764544"

   ],

API文档：

https://kubesphere.io/api/kubesphere/#tag/Pod-Metrics

中针对 metrics_filter 说明是：

The metric name filter consists of a regexp pattern. It specifies which metric data to return. For example, the following filter matches both pod CPU usage and memory usage: pod_cpu_usage|pod_memory_usage. View available metrics at kubesphere.io.

但如上，pod_memory_usage（4.2G）拿到的数值和web控制台中容器监控数值（1.7G）以及kubectl top （1.7G）并不一样，反而pod_memory_usage_wo_cache（1.7G）更接近，请问这是什么原因呢，我应该取哪个数值？