如何在Docker上将数据导入Elasticsearch？

Question

I made export from Elasticsearch on one machine and now want to import these data into another machine, where I have Elasticsearch running on docker.

This is the Elasticsearch-related content in docker-compose file:

elasticsearch:
    image: docker.elastic.co/elasticsearch/elasticsearch:5.5.0
    environment:
      - cluster.name=my-docker-cluster
      - bootstrap.memory_lock=true
      - "ES_JAVA_OPTS=-Xms512m -Xmx512m"
      - xpack.security.enabled=false
      - http.host=0.0.0.0
    ulimits:
      memlock:
        soft: -1
        hard: -1
      nofile:
        soft: 65536
        hard: 65536
    cap_add:
      - IPC_LOCK
    volumes:
      - ./src/elastic:/home
    ports:
      - 9200
      - 9300
    networks:
      - my_network

The elastic search container is up and running. However, I do not know how to import the data.

I found out the IP of Elasticsearch container (specified as XXX.XX.XX.X below) and executed the following command:

curl -X PUT -H "Content-Type: application/json" -H "Cache-Control: no-cache" -d '{
  "type": "fs",
  "settings": {
    "location": "/home/test/etk_backup/myproject_backup"
  }
}' "http://XXX.XX.XX.X:9200/_snapshot/myproject_backup"

But got this error:

{"error":{"root_cause":[{"type":"repository_exception","reason":"[myproject_backup] location [/home/test/etk_backup/myproject_backup] doesn't match any of the locations specified by path.repo because this setting is empty"}]

I entered into the container of Elasticsearch and manually specified path.repo in elasticsearch.yml as follows:

path-repo: ["/home/test/etk_backup/myproject_backup"]

But got the same error. How can I solve this issue?

Answer 1

尝试以下方法：

path.repo: ["/home/test/etk_backup/myproject_backup"]

Answer 2

To mention a convenient tool for this: we've been using Elasticdump to copy data from one index to another (typically for developers to have data in their local index). It's not clear if it's still maintained, but it still works for us on ES6.3

You would use it like so:

$ npm install elasticdump -g
$ elasticdump --input=https://XXX.XX.XX.X:9200/<index> --output /tmp/es-mapping --type=mapping --headers='{"Content-Type": "application/json"}'
$ elasticdump --input=https://XXX.XX.XX.X:9200/<index> --output /tmp/es-data --type=data --headers='{"Content-Type": "application/json"}' [--searchBody '<some query to filter what's being copied>']
$ elasticdump --input /tmp/es-mapping --output https://YYY.YY.YY.Y:9200/<index> --type mapping
$ elasticdump --input /tmp/es-data --output https://YYY.YY.YY.Y:9200/<index> --type data

You probably don't need the intermediate step of saving to a file, but I find it a nice thing to have in cases where you're experimenting locally and need to reload the index when your experiment failed in a destructive manner.