How to programmatically clean `data/transactions` folder in neo4j / auradb?

Neo4j Graph Platform Import / Export
, ,
1

Hello,

I'm doing an import of the data into neo4J and I'm running out of space because of the growing transaction folder.
With the database size close to 9GB I already have 45GB of transactions.

$ du -hs data/*
9.2G    data/databases
4.0K    data/dbms
0       data/dumps
4.0K    data/server_id
45G     data/transactions

Is it possible somehow to clean this folder or limit it's size using cypher (I'm asking about cypher as I'm planning to use auradb in the future)?

I use a list of queries like this one to import my data:

CALL apoc.periodic.iterate(
'
LOAD CSV WITH HEADERS FROM "https://path-to-the-csv-file.csv" AS row RETURN row
', '
    MERGE (n: MyNode {some_id: row.some_id})
    ON CREATE
        SET n.created = datetime(row.created),
               n.some_other_data = row.some_other_data
',
{batchSize:1000, parallel:false}) YIELD batches, total, operations
RETURN batches, total, operations;

Neo4j version: 5.6
Client: Neo4j Python Driver

Thank you for your help!

2

@dashkov.oleksandr

do not manually remove files.
However you can dynamically set txn log size and retention. See Update dynamic settings - Operations Manual

and logs are rotated based upon these settings during checkpoint which by default is every 15 mins or every 100txn. But you can also force via call db.checkpoint(); Oh but db.checkpoint(); is not available in Community. You can however change the default 15min / 100k txn via Configuration settings - Operations Manual and Configuration settings - Operations Manual

3

@dana_canzano , thank you for your answer.
I'm not sure if I'm doing everything correctly.

  1. I have default value of the txn log size set to 250M and retention to 15 min. A list of queries for execution takes more than an hour and as I've show before the folder growth until almost 45G. Why doesn't neo4j clean this folder?
  2. For the community version do you propose to increase db.checkpoint.interval.tx and db.checkpoint.interval.volume to have less transactions saved in this folder?

Thanks