Shell 脚本 - 基于“时间戳”属性对“AWS cloudwatch 指标”json 数组进行排序，原始输出包括统计信息

Question

I am running aws cli aws cloudwatch get-metric-statistics --metric-name CPUUtilization --start-time 2010-02-20T12:00:00 --end-time 2010-02-20T15:00:00 --period 60 --namespace AWS/EC2 --extended-statistics p80 --dimensions Name=InstanceId,Value=i-0b123423423

the output comes as

{
  "Label": "CPUUtilization",
  "Datapoints": [
    {
      "Timestamp": "2020-02-20T12:15:00Z",
      "Unit": "Percent",
      "ExtendedStatistics": {
        "p80": 0.16587132264856133
      }
    },

How do i get the output in the below format's (2 Columns)

19.514049550078127  2020-02-13T20:15:00Z
12.721997782508938  2020-02-13T19:15:00Z
13.318820949213313  2020-02-13T18:15:00Z
15.994192991030545  2020-02-13T17:15:00Z
18.13096421299414   2020-02-13T16:15:00Z

with Heading as CPUUtilization (2 columns)

CPUUtilization
19.514049550078127  2020-02-13T20:15:00Z
12.721997782508938  2020-02-13T19:15:00Z
13.318820949213313  2020-02-13T18:15:00Z
15.994192991030545  2020-02-13T17:15:00Z
18.13096421299414   2020-02-13T16:15:00Z

And in single column

19.514049550078127  
12.721997782508938  
13.318820949213313  
15.994192991030545  
18.13096421299414   

How can achieve this ?

Answer 1

Assuming the input file is input.json , then:

To output in the 2 columns format:

jq -r  '.Datapoints[] | [.ExtendedStatistics.p80, .Timestamp] | @tsv' input.json | sort -nr

With Heading as CPUUtilization (2 columns):

echo CPUUtilization; jq -r  '.Datapoints[] | [.ExtendedStatistics.p80, .Timestamp] | @tsv' input.json | sort -nr 

And in single column:

jq -r  '.Datapoints[] | [.ExtendedStatistics.p80] | @tsv' input.json | sort -nr