使用python生成大量数据写入es数据库并查询操作(2)

2024-04-02 19:55

短信预约 -IT技能 免费直播动态提醒

方案一

在写入数据时未提前创建索引mapping，而是每插入一条数据都包含了索引的信息。

示例代码：【多线程写入数据】【一次性写入10000*1000条数据】【本人亲测耗时3266秒】

from elasticsearch import Elasticsearch
from elasticsearch import helpers
from datetime import datetime
from queue import Queue
import random
import time
import threading
es = Elasticsearch(hosts='http://127.0.0.1:9200')
# print(es)
 
names = ['刘一', '陈二', '张三', '李四', '王五', '赵六', '孙七', '周八', '吴九', '郑十']
sexs = ['男', '女']
age = [25, 28, 29, 32, 31, 26, 27, 30]
character = ['自信但不自负,不以自我为中心',
             '努力、积极、乐观、拼搏是我的人生信条',
             '抗压能力强，能够快速适应周围环境',
             '敢做敢拼，脚踏实地；做事认真负责，责任心强',
             '爱好所学专业，乐于学习新知识；对工作有责任心；踏实，热情，对生活充满激情',
             '主动性强，自学能力强，具有团队合作意识，有一定组织能力',
             '忠实诚信,讲原则，说到做到，决不推卸责任',
             '有自制力，做事情始终坚持有始有终，从不半途而废',
             '肯学习,有问题不逃避,愿意虚心向他人学习',
             '愿意以谦虚态度赞扬接纳优越者,权威者',
             '会用100%的热情和精力投入到工作中；平易近人',
             '为人诚恳,性格开朗,积极进取,适应力强、勤奋好学、脚踏实地',
             '有较强的团队精神,工作积极进取,态度认真']
subjects = ['语文', '数学', '英语', '生物', '地理']
grades = [85, 77, 96, 74, 85, 69, 84, 59, 67, 69, 86, 96, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86]
create_time = datetime.now().strftime('%Y-%m-%d %H:%M:%S')
 
def save_to_es(num):
    """
    批量写入数据到es数据库
    :param num:
    :return:
    """
    start = time.time()
    action = [
        {
            "_index": "personal_info_10000000",
            "_type": "doc",
            "_id": i,
            "_source": {
                "id": i,
                "name": random.choice(names),
                "sex": random.choice(sexs),
                "age": random.choice(age),
                "character": random.choice(character),
                "subject": random.choice(subjects),
                "grade": random.choice(grades),
                "create_time": create_time
            }
        } for i in range(10000 * num, 10000 * num + 10000)
    ]
    helpers.bulk(es, action)
    end = time.time()
    print(f"{num}耗时{end - start}s!")
 
def run():
    global queue
    while queue.qsize() > 0:
        num = queue.get()
        print(num)
        save_to_es(num)

if __name__ == '__main__':
    start = time.time()
    queue = Queue()
    # 序号数据进队列
    for num in range(1000):
        queue.put(num)
 
    # 多线程执行程序
    consumer_lst = []
    for _ in range(10):
        thread = threading.Thread(target=run)
        thread.start()
        consumer_lst.append(thread)
    for consumer in consumer_lst:
        consumer.join()
    end = time.time()
    print('程序执行完毕！花费时间：', end - start)

运行结果：

自动创建的索引mapping：

GET personal_info_10000000/_mapping
{
  "personal_info_10000000" : {
    "mappings" : {
      "properties" : {
        "age" : {
          "type" : "long"
        },
        "character" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 256
            }
          }
        },
        "create_time" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 256
            }
          }
        },
        "grade" : {
          "type" : "long"
        },
        "id" : {
          "type" : "long"
        },
        "name" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 256
            }
          }
        },
        "sex" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 256
            }
          }
        },
        "subject" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 256
            }
          }
        }
      }
    }
  }
}

方案二

1.顺序插入5000000条数据

先创建索引personal_info_5000000，确定好mapping后，再插入数据。

新建索引并设置mapping信息：

PUT personal_info_5000000
{
  "settings": {
    "number_of_shards": 3,
    "number_of_replicas": 1
  },
  "mappings": {
    "properties": {
      "id": {
        "type": "long"
      },
      "name": {
        "type": "text",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 32
          }
        }
      },
      "sex": {
        "type": "text",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 8
          }
        }
      },
      "age": {
        "type": "long"
      },
      "character": {
        "type": "text",
        "analyzer": "ik_smart",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 256
          }
        }
      },
      "subject": {
        "type": "text",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 256
          }
        }
      },
      "grade": {
        "type": "long"
      },
      "create_time": {
        "type": "date",
        "format": "yyyy-MM-dd HH:mm:ss||yyyy-MM-dd||epoch_millis"
      }
    }
  }
}

查看新建索引信息：

GET personal_info_5000000
 
{
  "personal_info_5000000" : {
    "aliases" : { },
    "mappings" : {
      "properties" : {
        "age" : {
          "type" : "long"
        },
        "character" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 256
            }
          },
          "analyzer" : "ik_smart"
        },
        "create_time" : {
          "type" : "date",
          "format" : "yyyy-MM-dd HH:mm:ss||yyyy-MM-dd||epoch_millis"
        },
        "grade" : {
          "type" : "long"
        },
        "id" : {
          "type" : "long"
        },
        "name" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 32
            }
          }
        },
        "sex" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 8
            }
          }
        },
        "subject" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 256
            }
          }
        }
      }
    },
    "settings" : {
      "index" : {
        "routing" : {
          "allocation" : {
            "include" : {
              "_tier_preference" : "data_content"
            }
          }
        },
        "number_of_shards" : "3",
        "provided_name" : "personal_info_50000000",
        "creation_date" : "1663471072176",
        "number_of_replicas" : "1",
        "uuid" : "5DfmfUhUTJeGk1k4XnN-lQ",
        "version" : {
          "created" : "7170699"
        }
      }
    }
  }
}

开始插入数据：

示例代码： 【单线程写入数据】【一次性写入10000*500条数据】【本人亲测耗时7916秒】

from elasticsearch import Elasticsearch
from datetime import datetime
from queue import Queue
import random
import time
import threading
es = Elasticsearch(hosts='http://127.0.0.1:9200')
# print(es)
names = ['刘一', '陈二', '张三', '李四', '王五', '赵六', '孙七', '周八', '吴九', '郑十']
sexs = ['男', '女']
age = [25, 28, 29, 32, 31, 26, 27, 30]
character = ['自信但不自负,不以自我为中心',
             '努力、积极、乐观、拼搏是我的人生信条',
             '抗压能力强，能够快速适应周围环境',
             '敢做敢拼，脚踏实地；做事认真负责，责任心强',
             '爱好所学专业，乐于学习新知识；对工作有责任心；踏实，热情，对生活充满激情',
             '主动性强，自学能力强，具有团队合作意识，有一定组织能力',
             '忠实诚信,讲原则，说到做到，决不推卸责任',
             '有自制力，做事情始终坚持有始有终，从不半途而废',
             '肯学习,有问题不逃避,愿意虚心向他人学习',
             '愿意以谦虚态度赞扬接纳优越者,权威者',
             '会用100%的热情和精力投入到工作中；平易近人',
             '为人诚恳,性格开朗,积极进取,适应力强、勤奋好学、脚踏实地',
             '有较强的团队精神,工作积极进取,态度认真']
subjects = ['语文', '数学', '英语', '生物', '地理']
grades = [85, 77, 96, 74, 85, 69, 84, 59, 67, 69, 86, 96, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86]
create_time = datetime.now().strftime('%Y-%m-%d %H:%M:%S')
 
# 添加程序耗时的功能
def timer(func):
    def wrapper(*args, **kwargs):
        start = time.time()
        res = func(*args, **kwargs)
        end = time.time()
        print('id{}共耗时约 {:.2f} 秒'.format(*args, end - start))
        return res
 
    return wrapper
 
@timer
def save_to_es(num):
    """
    顺序写入数据到es数据库
    :param num:
    :return:
    """
    body = {
        "id": num,
        "name": random.choice(names),
        "sex": random.choice(sexs),
        "age": random.choice(age),
        "character": random.choice(character),
        "subject": random.choice(subjects),
        "grade": random.choice(grades),
        "create_time": create_time
    }
    # 此时若索引不存在时会新建
    es.index(index="personal_info_5000000", id=num, doc_type="_doc", document=body)
 
def run():
    global queue
    while queue.qsize() > 0:
        num = queue.get()
        print(num)
        save_to_es(num)
 
if __name__ == '__main__':
    start = time.time()
    queue = Queue()
    # 序号数据进队列
    for num in range(5000000):
        queue.put(num)
 
    # 多线程执行程序
    consumer_lst = []
    for _ in range(10):
        thread = threading.Thread(target=run)
        thread.start()
        consumer_lst.append(thread)
    for consumer in consumer_lst:
        consumer.join()
    end = time.time()
    print('程序执行完毕！花费时间：', end - start)

运行结果：

2.批量插入5000000条数据

先创建索引personal_info_5000000_v2，确定好mapping后，再插入数据。

新建索引并设置mapping信息：

PUT personal_info_5000000_v2
{
  "settings": {
    "number_of_shards": 3,
    "number_of_replicas": 1
  },
  "mappings": {
    "properties": {
      "id": {
        "type": "long"
      },
      "name": {
        "type": "text",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 32
          }
        }
      },
      "sex": {
        "type": "text",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 8
          }
        }
      },
      "age": {
        "type": "long"
      },
      "character": {
        "type": "text",
        "analyzer": "ik_smart",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 256
          }
        }
      },
      "subject": {
        "type": "text",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 256
          }
        }
      },
      "grade": {
        "type": "long"
      },
      "create_time": {
        "type": "date",
        "format": "yyyy-MM-dd HH:mm:ss||yyyy-MM-dd||epoch_millis"
      }
    }
  }
}

查看新建索引信息：

GET personal_info_5000000_v2
 
{
  "personal_info_5000000_v2" : {
    "aliases" : { },
    "mappings" : {
      "properties" : {
        "age" : {
          "type" : "long"
        },
        "character" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 256
            }
          },
          "analyzer" : "ik_smart"
        },
        "create_time" : {
          "type" : "date",
          "format" : "yyyy-MM-dd HH:mm:ss||yyyy-MM-dd||epoch_millis"
        },
        "grade" : {
          "type" : "long"
        },
        "id" : {
          "type" : "long"
        },
        "name" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 32
            }
          }
        },
        "sex" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 8
            }
          }
        },
        "subject" : {
          "type" : "text",
          "fields" : {
            "keyword" : {
              "type" : "keyword",
              "ignore_above" : 256
            }
          }
        }
      }
    },
    "settings" : {
      "index" : {
        "routing" : {
          "allocation" : {
            "include" : {
              "_tier_preference" : "data_content"
            }
          }
        },
        "number_of_shards" : "3",
        "provided_name" : "personal_info_5000000_v2",
        "creation_date" : "1663485323617",
        "number_of_replicas" : "1",
        "uuid" : "XBPaDn_gREmAoJmdRyBMAA",
        "version" : {
          "created" : "7170699"
        }
      }
    }
  }
}

批量插入数据：

通过elasticsearch模块导入helper，通过helper.bulk来批量处理大量的数据。首先将所有的数据定义成字典形式，各字段含义如下：

_index对应索引名称，并且该索引必须存在。
_type对应类型名称。
_source对应的字典内，每一篇文档的字段和值，可有有多个字段。

示例代码： 【程序中途异常，写入4714000条数据】

from elasticsearch import Elasticsearch
from elasticsearch import helpers
from datetime import datetime
from queue import Queue
import random
import time
import threading
es = Elasticsearch(hosts='http://127.0.0.1:9200')
# print(es)
names = ['刘一', '陈二', '张三', '李四', '王五', '赵六', '孙七', '周八', '吴九', '郑十']
sexs = ['男', '女']
age = [25, 28, 29, 32, 31, 26, 27, 30]
character = ['自信但不自负,不以自我为中心',
             '努力、积极、乐观、拼搏是我的人生信条',
             '抗压能力强，能够快速适应周围环境',
             '敢做敢拼，脚踏实地；做事认真负责，责任心强',
             '爱好所学专业，乐于学习新知识；对工作有责任心；踏实，热情，对生活充满激情',
             '主动性强，自学能力强，具有团队合作意识，有一定组织能力',
             '忠实诚信,讲原则，说到做到，决不推卸责任',
             '有自制力，做事情始终坚持有始有终，从不半途而废',
             '肯学习,有问题不逃避,愿意虚心向他人学习',
             '愿意以谦虚态度赞扬接纳优越者,权威者',
             '会用100%的热情和精力投入到工作中；平易近人',
             '为人诚恳,性格开朗,积极进取,适应力强、勤奋好学、脚踏实地',
             '有较强的团队精神,工作积极进取,态度认真']
subjects = ['语文', '数学', '英语', '生物', '地理']
grades = [85, 77, 96, 74, 85, 69, 84, 59, 67, 69, 86, 96, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86]
create_time = datetime.now().strftime('%Y-%m-%d %H:%M:%S')
# 添加程序耗时的功能
def timer(func):
    def wrapper(*args, **kwargs):
        start = time.time()
        res = func(*args, **kwargs)
        end = time.time()
        print('id{}共耗时约 {:.2f} 秒'.format(*args, end - start))
        return res
 
    return wrapper
 
 
@timer
def save_to_es(num):
    """
    批量写入数据到es数据库
    :param num:
    :return:
    """
    action = [
        {
            "_index": "personal_info_5000000_v2",
            "_type": "_doc",
            "_id": i,
            "_source": {
                "id": i,
                "name": random.choice(names),
                "sex": random.choice(sexs),
                "age": random.choice(age),
                "character": random.choice(character),
                "subject": random.choice(subjects),
                "grade": random.choice(grades),
                "create_time": create_time
            }
        } for i in range(10000 * num, 10000 * num + 10000)
    ]
    helpers.bulk(es, action)
def run():
    global queue
    while queue.qsize() > 0:
        num = queue.get()
        print(num)
        save_to_es(num)
if __name__ == '__main__':
    start = time.time()
    queue = Queue()
    # 序号数据进队列
    for num in range(500):
        queue.put(num)
 
    # 多线程执行程序
    consumer_lst = []
    for _ in range(10):
        thread = threading.Thread(target=run)
        thread.start()
        consumer_lst.append(thread)
    for consumer in consumer_lst:
        consumer.join()
    end = time.time()
    print('程序执行完毕！花费时间：', end - start)

运行结果：

3.批量插入50000000条数据

先创建索引personal_info_5000000_v2，确定好mapping后，再插入数据。

此过程是在上面批量插入的前提下进行优化，采用python生成器。

建立索引和mapping同上，直接上代码：

示例代码： 【程序中途异常，写入3688000条数据】

from elasticsearch import Elasticsearch
from elasticsearch import helpers
from datetime import datetime
from queue import Queue
import random
import time
import threading
es = Elasticsearch(hosts='http://127.0.0.1:9200')
# print(es)
 
names = ['刘一', '陈二', '张三', '李四', '王五', '赵六', '孙七', '周八', '吴九', '郑十']
sexs = ['男', '女']
age = [25, 28, 29, 32, 31, 26, 27, 30]
character = ['自信但不自负,不以自我为中心',
             '努力、积极、乐观、拼搏是我的人生信条',
             '抗压能力强，能够快速适应周围环境',
             '敢做敢拼，脚踏实地；做事认真负责，责任心强',
             '爱好所学专业，乐于学习新知识；对工作有责任心；踏实，热情，对生活充满激情',
             '主动性强，自学能力强，具有团队合作意识，有一定组织能力',
             '忠实诚信,讲原则，说到做到，决不推卸责任',
             '有自制力，做事情始终坚持有始有终，从不半途而废',
             '肯学习,有问题不逃避,愿意虚心向他人学习',
             '愿意以谦虚态度赞扬接纳优越者,权威者',
             '会用100%的热情和精力投入到工作中；平易近人',
             '为人诚恳,性格开朗,积极进取,适应力强、勤奋好学、脚踏实地',
             '有较强的团队精神,工作积极进取,态度认真']
subjects = ['语文', '数学', '英语', '生物', '地理']
grades = [85, 77, 96, 74, 85, 69, 84, 59, 67, 69, 86, 96, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86]
create_time = datetime.now().strftime('%Y-%m-%d %H:%M:%S')
 
# 添加程序耗时的功能
def timer(func):
    def wrapper(*args, **kwargs):
        start = time.time()
        res = func(*args, **kwargs)
        end = time.time()
        print('id{}共耗时约 {:.2f} 秒'.format(*args, end - start))
        return res
 
    return wrapper
@timer
def save_to_es(num):
    """
    使用生成器批量写入数据到es数据库
    :param num:
    :return:
    """
    action = (
        {
            "_index": "personal_info_5000000_v3",
            "_type": "_doc",
            "_id": i,
            "_source": {
                "id": i,
                "name": random.choice(names),
                "sex": random.choice(sexs),
                "age": random.choice(age),
                "character": random.choice(character),
                "subject": random.choice(subjects),
                "grade": random.choice(grades),
                "create_time": create_time
            }
        } for i in range(10000 * num, 10000 * num + 10000)
    )
    helpers.bulk(es, action)
 
def run():
    global queue
    while queue.qsize() > 0:
        num = queue.get()
        print(num)
        save_to_es(num)
 
if __name__ == '__main__':
    start = time.time()
    queue = Queue()
    # 序号数据进队列
    for num in range(500):
        queue.put(num)
 
    # 多线程执行程序
    consumer_lst = []
    for _ in range(10):
        thread = threading.Thread(target=run)
        thread.start()
        consumer_lst.append(thread)
    for consumer in consumer_lst:
        consumer.join()
    end = time.time()
    print('程序执行完毕！花费时间：', end - start)

运行结果：