28 Jan 2020

Neo4j: Performing a database dump within a Docker container

Before the release of Neo4j 4.0, taking a dump of a database running within a Docker container was a tricky affair.

We’d need to stop the container and remove it, run the container again in bash mode, and finally take a dump of the database. With 4.0 things are simpler.

Figure 1. Neo4j on Docker

We’ll be using the following Docker Compose configuration in this blog post:

Dockerfile

  version: '3.7'

  services:
    neo4j:
      image: neo4j:4.0.0-enterprise
      container_name: "quickgraph-aus-open"
      volumes:
        - ./plugins:/plugins
        - ./data:/data
        - ./import:/var/lib/neo4j/import
      ports:
        - "7474:7474"
        - "7687:7687"
      environment:
        - "NEO4J_ACCEPT_LICENSE_AGREEMENT=yes"
        - "NEO4J_AUTH=neo4j/neo"

Once we’ve created that file we need to open a terminal session where that file lives and then run docker-compose up to spin up Neo4j. When that’s finished running, we can then run the following command to check on the status of our container:

$ docker ps

CONTAINER ID        IMAGE                    COMMAND                  CREATED             STATUS              PORTS                                                      NAMES
77b91b30aa88        neo4j:4.0.0-enterprise   "/sbin/tini -g -- /d…"   8 seconds ago       Up 6 seconds        0.0.0.0:7474->7474/tcp, 7473/tcp, 0.0.0.0:7687->7687/tcp   quickgraph-aus-open

Let’s check which databases our Neo4j instance contains by running the SHOW DATABASES command:

$ docker exec -it quickgraph-aus-open cypher-shell -u neo4j -p neo -d system "SHOW DATABASES;"

Table 1. SHOW DATABASES
name	address	role	requestedStatus	currentStatus	error	default
"neo4j"	"0.0.0.0:7687"	"standalone"	"online"	"online"	""	TRUE
"system"	"0.0.0.0:7687"	"standalone"	"online"	"online"	""	FALSE
"womens"	"0.0.0.0:7687"	"standalone"	"online"	"online"	""	FALSE

We’re going to take a dump of the womens database using the Neo4j Admin dump command. We can call that command by running the following:

$ docker exec -it quickgraph-aus-open neo4j-admin dump --database=womens --to=/tmp/womens-aus-open.dump
The database is in use. Stop database 'womens' and try again.

Let’s stop that database using the STOP DATABASE command:

$ docker exec -it quickgraph-aus-open cypher-shell -u neo4j -p neo -d system "STOP DATABASE womens;"
0 rows available after 268 ms, consumed after another 0 ms

And now let’s re-run the dump command:

$ docker exec -it quickgraph-aus-open neo4j-admin dump --database=womens --to=/tmp/womens-aus-open.dump
Done: 75 files, 252.9MiB processed.

The dump file is written to the Docker container, so now we need to copy that file onto the host:

docker cp quickgraph-aus-open:/tmp/womens-aus-open.dump .

And finally let’s start our database back up again:

$ docker exec -it quickgraph-aus-open cypher-shell -u neo4j -p neo -d system "START DATABASE womens;"
0 rows available after 238 ms, consumed after another 0 ms

And we’re done!

About the author

I'm currently working on short form content at ClickHouse. I publish short 5 minute videos showing how to solve data problems on YouTube @LearnDataWithMark. I previously worked on graph analytics at Neo4j, where I also co-authored the O'Reilly Graph Algorithms Book with Amy Hodler.