elasticsearch python parallel_bulk无法插入数据

Question

我曾经parallel_bulk在python中向Elasticsearch插入数据，但parallel_bulk无法插入数据。我的代码：

class CreateIndex(object):
def _gen_data(self, index, doc_type, chunk_size):
    sql = """select * from tem_search_engine_1 where rownum <= 10000"""  
    self.cursor.execute(sql)
    col_name_list = [col[0].lower() for col in self.cursor.description]
    col_name_len = len(col_name_list)
    actions = []

    start = time.time()
    for row in self.cursor:
        source = {}
        tbl_id = ""
        for i in range(col_name_len):
            source.update({col_name_list[i]: str(row[i])})
            if col_name_list[i] == "tbl_id":
                tbl_id = row[i]
        action = {
            "_index": index,
            "_type": doc_type,
            "_id": tbl_id,  
            "_source": source
        }
        actions.append(action)
        if len(actions) == chunk_size:
            print("actions time：", time.time()-start)
            yield actions
            actions = []
    print("for time：", time.time()-start)
    yield actions

def bulk_data(self, index, doc_type, chunk_size=1000, is_parallel=True, threads_counts=4):

    t1 = time.time()
    gen_action = self._gen_data(index, doc_type, chunk_size)

    if is_parallel is None or is_parallel == True:
        for success, info in helpers.parallel_bulk(client=self.es, actions=gen_action, thread_count=threads_counts):
            if not success:
                print("Insert failed: ", info)

if __name__ == "__main__":
    createindex = CreateIndex()
    createindex.create_index(index="se", doc_type="se_doc")
    createindex.bulk_data(index="se", doc_type="se_doc")

当我使用bulk_data，但无法插入任何数据时，该如何处理呢？错误是：

Traceback (most recent call last):
  File "F:/programs/ElasticSearch/CreateIndex.py", line 287, in <module>
    createindex.bulk_data(index="se", doc_type="se_doc")
  File "F:/programs/ElasticSearch/CreateIndex.py", line 179, in bulk_data
    thread_count=threads_counts, chunk_size=chunk_size):
  File "F:\programs\ElasticSearch\lib\site-packages\elasticsearch\helpers\__init__.py", line 306, in parallel_bulk
    _chunk_actions(actions, chunk_size, max_chunk_bytes, client.transport.serializer)
  File "D:\anacond\lib\multiprocessing\pool.py", line 735, in next
    raise value
  File "D:\anacond\lib\multiprocessing\pool.py", line 119, in worker
    result = (True, func(*args, **kwds))
  File "D:\anacond\lib\multiprocessing\pool.py", line 138, in _helper_reraises_exception
    raise ex
  File "D:\anacond\lib\multiprocessing\pool.py", line 290, in _guarded_task_generation
    for i, x in enumerate(iterable):
  File "F:\programs\ElasticSearch\lib\site-packages\elasticsearch\helpers\__init__.py", line 58, in _chunk_actions
    for action, data in actions:
  File "F:\programs\ElasticSearch\lib\site-packages\elasticsearch\helpers\__init__.py", line 37, in expand_action
    op_type = data.pop('_op_type', 'index')
TypeError: pop() takes at most 1 argument (2 given)

Answer 1

使用parallel_bulkmethod，您可以传递一个字典列表或一个生成器，生成一个字典。在这里解释。一个generator在python用于未在RAM中加载变量，但如果你要通过你的elem前一个清单-
在字典action中的列表actions，它没有更多的意义，因为建立一个列表，你应该在内存中加载所有里面的元素。在您的情况下，您要传递的生成器不会产生dict符action--但会生成操作列表-
actions。

因此，或者您的函数_gen_data返回一个列表，实际上是一个生成器的列表：

def _gen_data(self, index, doc_type, chunk_size):
    sql = """select * from tem_search_engine_1 where rownum <= 10000"""  
    self.cursor.execute(sql)
    col_name_list = [col[0].lower() for col in self.cursor.description]
    col_name_len = len(col_name_list)
    actions = []

    start = time.time()
    for row in self.cursor:
        source = {}
        tbl_id = ""
        for i in range(col_name_len):
            source.update({col_name_list[i]: str(row[i])})
            if col_name_list[i] == "tbl_id":
                tbl_id = row[i]
        action = {
            "_index": index,
            "_type": doc_type,
            "_id": tbl_id,  
            "_source": source
        }
        actions.append(action)
    return actions

或者，您不创建actions列表，而得出action字典：

def _gen_data(self, index, doc_type, chunk_size):
    sql = """select * from tem_search_engine_1 where rownum <= 10000"""  
    self.cursor.execute(sql)
    col_name_list = [col[0].lower() for col in self.cursor.description]
    col_name_len = len(col_name_list)


    start = time.time()
    for row in self.cursor:
        source = {}
        tbl_id = ""
        for i in range(col_name_len):
            source.update({col_name_list[i]: str(row[i])})
            if col_name_list[i] == "tbl_id":
                tbl_id = row[i]
        yield {
            "_index": index,
            "_type": doc_type,
            "_id": tbl_id,  
            "_source": source
        }