Skip to content

SYSTEM Cited by 1 source

ArangoDB

ArangoDB is an open-source multi-model database combining document, graph, and key-value data models under a single query language (AQL — ArangoDB Query Language). Originally released by ArangoDB GmbH (Cologne, Germany) in 2012. Supports geospatial indexing and queries as a built-in capability.

Within the wiki this page is a stub created for cross-referencing from sources/2025-10-12-mongodb-cars24-improves-search-for-300-million-users-with-atlas — Cars24 ran a geospatial-search workload on ArangoDB before migrating to MongoDB Atlas.

Role in the wiki

Legacy substrate in the Cars24 geospatial-search migration

Cars24 used ArangoDB for geospatial search before consolidating onto MongoDB Atlas. MongoDB's case study (2025-10-12) names three architectural limitations that drove the move:

Cars24-named pain Description
Performance bottlenecks Not quantified in the post.
Weak transactions "Difficult to guarantee consistent data operations" — Cars24 framing.
Limited ecosystem "Scaling developer onboarding and troubleshooting became increasingly onerous" — ecosystem / community depth as a scaling constraint on a fast-hiring engineering org.

Post-migration outcome reported: ~50 % cost reduction moving to MongoDB Atlas; robust multi-document ACID transactions across shards; consolidated operations under one platform.

(Source: sources/2025-10-12-mongodb-cars24-improves-search-for-300-million-users-with-atlas)

Framing caveat

This is MongoDB-authored case-study framing — one vendor characterizing a competitor's limitations from a migrated-customer perspective. ArangoDB does support multi-document transactions (introduced in ArangoDB 3.7, 2020) and is actively maintained. The "weak transactions" label is Cars24's framing of what fit their workload, not a neutral technical assessment. Treat as: "ArangoDB's transaction model + ecosystem + performance shape did not fit Cars24's scaling trajectory", not "ArangoDB is transactionally weak".

Stub caveats

Not covered on this page (the sysdesign-wiki doesn't have a deep ArangoDB source to extract from):

  • AQL query language internals.
  • ArangoDB's graph / document / key-value unified storage layer.
  • ArangoSearch (ArangoDB's full-text search engine based on iresearch).
  • Cluster architecture, sharding model, failover behaviour.
  • ArangoDB Oasis (managed cloud offering).
  • 3.7+ multi-document transaction semantics + guarantees.

Seen in

Last updated · 200 distilled / 1,178 read