nebula-python 无法顺序执行,有时候创建了tag,有时候没创建tag,无法创建数据,必须多次运行才能全部创建达到理想效果,请问这是什么问题

import json
from nebula3.gclient.net import ConnectionPool
from nebula3.Config import Config

def uploadJson():
    # 从包含嵌套JSON数据的文件中读取数据
    with open('data.json') as file:
        data = json.load(file)
    # 遍历JSON数据
    for key in data:
        if key['tag'] == "player":
            sqlCreateTag = 'CREATE TAG IF NOT EXISTS `player`(`name` string, `age` int)'
        elif key['tag'] == "team":
            sqlCreateTag = 'CREATE TAG IF NOT EXISTS `team`(`name` string)'
        print(sqlCreateTag)
        print(session.execute(sqlCreateTag))
        propsName = '('
        propsData = '('
        # 遍历JSON数据中的每个元素
        for item in key:
            if item == "tag" or item == "id":
                continue
            if type(key[item]) == str:
                propsData += ('"' + key[item] + '", ')
                propsName += (item + ', ')
            else:
                propsData += (str(key[item]) + ', ')
                propsName += (item + ', ')
        propsName = propsName[:len(propsName) - 2]
        propsData = propsData[:len(propsData) - 2]
        propsName += ')'
        propsData += ')'
        sqlInsertVertex = 'INSERT VERTEX {} {} VALUES "{}":{}'.format(key['tag'], propsName, key['id'], propsData)
        print(session.execute(sqlInsertVertex))
        print(sqlInsertVertex)


if __name__ == '__main__':
    # 定义配置
    config = Config()
    config.max_connection_pool_size = 10
    config.timeout = 500
    # 初始化连接池
    connection_pool = ConnectionPool()
    # 如果给定的服务器正常,则返回true,否则返回false。
    ok = connection_pool.init([('192.168.161.128', 9669)], config)
    # 方法1:控制连接自行释放。
    # 从连接池中获取会话
    session = connection_pool.get_session('root', 'nebula')
    # 创建Test空间
    session.execute("CREATE SPACE IF NOT EXISTS `Test` (vid_type = FIXED_STRING(32))")
    # 使用Test空间
    session.execute("USE Test")
    uploadJson()
    # 释放会话
    session.release()
    # 关闭连接池
    connection_pool.close()

修改完 schema 后,建议 sleep 20s(即两个心跳),可以确保 schema 肯定生效(背后原因是因为有异步流程)