现在留言不是实时发布了吗？ 刚刚留言说有问题，现在原因找到了，多加个一个size: 0,。但是发现这个聚合很没有意思，用下面简单的query就能达到同样效果。
# query
GET employees/_search
{
  "query": {
    "match": {
      "job.keyword": "Dev Manager"
    }
  }
}
# result
{
  "took" : 2,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 2,
      "relation" : "eq"
    },
    "max_score" : 2.1747518,
    "hits" : [
      {
        "_index" : "employees",
        "_type" : "_doc",
        "_id" : "2",
        "_score" : 2.1747518,
        "_source" : {
          "name" : "Underwood",
          "age" : 41,
          "job" : "Dev Manager",
          "gender" : "male",
          "salary" : 50000
        }
      },
      {
        "_index" : "employees",
        "_type" : "_doc",
        "_id" : "21",
        "_score" : 2.1747518,
        "_source" : {
          "name" : "Underwood2",
          "age" : 42,
          "job" : "Dev Manager",
          "gender" : "male",
          "salary" : 60000
        }
      }
    ]
  }
}
我期望达到的效果是：结果中只有"Dev Manager"job类型的，按terms分桶的数据。
比如：
{
   "key" : "Dev Manager",
   "doc_count" : 1
}
这个才是我期望的聚合后再对结果进行post_filter的效果，相当于对结果集再做处理，但不知道ES是不是这样设计的。或者应该用pipeline达到我想要的结果。

Post_filter那个例子没达到预期的结果啊？并没有把Dev Manager作为一个单独的桶返回。从Query的Result看，post_filter 没起作用。
# query:
POST employees/_search
{
  "size": 0,
  "aggs": {
    "job_cnt": {
      "terms": {
        "field": "job.keyword"
      }
    }
  },
  "post_filter": {
    "match": {
      "job.keyword" : "Dev Manager"
    }
  }
}

#Result
{
  "took" : 1,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 0,
      "relation" : "eq"
    },
    "max_score" : null,
    "hits" : [ ]
  },
  "aggregations" : {
    "job_cnt" : {
      "doc_count_error_upper_bound" : 0,
      "sum_other_doc_count" : 0,
      "buckets" : [
        {
          "key" : "Java Programmer",
          "doc_count" : 7
        },
        {
          "key" : "Javascript Programmer",
          "doc_count" : 4
        },
        {
          "key" : "QA",
          "doc_count" : 3
        },
        {
          "key" : "DBA",
          "doc_count" : 2
        },
        {
          "key" : "Web Designer",
          "doc_count" : 2
        },
        {
          "key" : "Dev Manager",
          "doc_count" : 1
        },
        {
          "key" : "Product Manager",
          "doc_count" : 1
        }
      ]
    }
  }
}