How to compact the growing collections?

[ohsp123-arango]

trafficstars

My Environment

• ArangoDB Version: Enterprise 3.7.2
• Storage Engine: RocksDB
• Deployment Mode: Cluster
• Deployment Strategy: ArangoDB Starter
• Configuration: 3 nodes cluster
• Infrastructure: own servers
• Operating System: RHEL 8.2
• Total RAM in your machine: 64GB each node
• Disks in use: 420GB on each node
• Used Package: RedHat .rpm

Component, Query & Data

Affected feature:

AQL query (if applicable):

AQL explain and/or profile (if applicable):

Dataset:

Size of your Dataset on disk:

400GB

Replication Factor & Number of Shards (Cluster only):

Replication factor=2

Steps to reproduce

Problem: the size of my collection is growing and cannot reclaim the disk spaces and it's now almost 90% full

Expected result: need to reclaim the disk spaces after deleting collections from application (Prophet Enterprise)

[dothebart]

Hi, please note that ArangoDB 3.7.13 is available for download. It contains lots of bugfixes above your current version (3.7.2 if I get that correctly?) and one of them is to be able to retain disk space.

Creating and truncating a collection could probably also trigger this, but it could be that this was only added in a later release.

[Simran-B]

You can also trigger compaction manually if necessary: https://www.arangodb.com/docs/stable/data-modeling-databases-working-with.html#compact

[ohsp123-arango]

Hello,

I've tried to compact with db._compact() command but get the answer, only allowed for superusers. What is superuser? I tried it with ArangoDB root account and didin't know what is 'superuser' means? Please advise me..

[pashkasan]

have the same issue on 3.8.4 cluster