Star

importer导入csv失败, No valid connection in the idle queue and connection number has reached the pool capacity

为了更快地定位、解决问题,麻烦参考下面模版提问 ^ ^

提问参考模版:

  • nebula 版本:1.2.1-beta
  • 部署方式(分布式 / Docker ):docker swarm部署虚拟机上的两个节点
  • 硬件信息
    • 磁盘( 必须为 SSD ,不支持 HDD)
    • CPU、内存信息:
  • 出问题的 Space 的创建方式:执行 describe space xxx;
  • 问题的具体描述

我全程follow这一篇教程 https://docs.nebula-graph.com.cn/manual-CN/1.overview/2.quick-start/4.import-csv-file/
但是没有导入成功。以下是报错和config.yaml

2020/12/17 12:53:16 --- START OF NEBULA IMPORTER ---
2020/12/17 12:53:16 [WARN] config.go:217: Invalid retry option in clientSettings.retry, reset to 1 
2020/12/17 12:53:16 [WARN] config.go:168: You have not configured whether to remove generated temporary files, reset to default value. removeTempFiles: false
2020/12/17 12:53:16 [INFO] connection_pool.go:74: [nebula-clients] connection pool is initialized successfully
2020/12/17 13:23:37 Failed to get connection: No valid connection in the idle queue and connection number has reached the pool capacity
2020/12/17 13:23:38 --- END OF NEBULA IMPORTER ---

version: v1
description: example
clientSettings:
  concurrency: 2 # number of graph clients
  channelBufferSize: 50
  space: nba
  connection:
    user: user
    password: password
    address: 172.16.9.129:3699,172.16.9.130:3699
logPath: ./err/test.log
files:
  - path: /home/jiayizhou/nebula/serve.csv
    failDataPath: ./err/serve.csv
    batchSize: 10
    type: csv
    csv:
      withHeader: false
      withLabel: false
    schema:
      type: edge
      edge:
        name: serve
        withRanking: false
        props:
          - name: start_year
            type: int
          - name: end_year
            type: int
  - path: /home/jiayizhou/nebula/follow.csv
    failDataPath: ./err/follow.csv
    batchSize: 10
    type: csv
    csv:
      withHeader: false
      withLabel: false
    schema:
      type: edge
      edge:
        name: follow
        withRanking: false
        props:
          - name: degree
            type: int
  - path: /home/jiayizhou/nebula/player.csv
    failDataPath: ./err/player.csv
    batchSize: 10
    type: csv
    csv:
      withHeader: false
      withLabel: false
    schema:
      type: vertex
      vertex:
        tags:
          - name: player
            props:
              - name: name
                type: string
              - name: age
                type: int
  - path: /home/jiayizhou/nebula/team.csv
    failDataPath: ./err/team.csv
    batchSize: 10
    type: csv
    csv:
      withHeader: false
      withLabel: false
    schema:
      type: vertex
      vertex:
        tags:
          - name: team
            props:
              - name: name
                type: string

你好,当前nebula-importer的master分支是2.x版本的, 1.x版本的在v1分支上。

nebula没有nebula 版本:1.2.1-beta 这个版本呀, 你使用的是哪个版本的?

我使用的是docker方式来导入csv
$ docker run --rm -ti --network=host
-v /home/nebula/config.yaml:/home/nebula/config.yaml
-v /home/nebula/:/home/nebula/ vesoft/nebula-importer
–config /home/nebula/config.yaml

本地有nebula-importer镜像:
vesoft/nebula-importer latest 269ae23d2a5f 3 days ago 15.1MB
请问这个是importer v1还是v2呢

我的nebula graph是通过docker swarm部署集群的方式搭建的(https://docs.nebula-graph.com.cn/manual-CN/3.build-develop-and-administration/2.install/deploy-nebula-with-swarm/)
studio右上角显示的是v1.2.1-beta

这个应该是2.0 nebula-importer

nebula-studio v1.2.1-beta对应的nebula-graph是1.x版本的,建议你使用nebula-importer:v1镜像再试一次。

我pull了v1镜像后出现了新的error:space not found for ‘nba’

但是我在studio和console里是能看到nba这个space的:
Screen Shot 2020-12-18 at 10.24.17 AM

浙ICP备20010487号