Project Voldemort
discontinuedby LinkedIn · since 2009 · also known as Voldemort
LinkedIn's discontinued open-source Dynamo clone, retired in favour of Venice.
discontinued. No longer developed; LinkedIn ended production use in 2018 and migrated to Venice. Last release 1.10.26 (2017).
Overview
Voldemort was LinkedIn's distributed key-value store, one of the first open-source implementations of Amazon's Dynamo paper, with consistent hashing, vector-clock versioning, pluggable storage engines and a well-known read-only mode for bulk-loading Hadoop output. It served LinkedIn features such as 'People You May Know' for years. LinkedIn stopped all production use in 2018 and the README states the project is no longer under development; the read-only use cases moved to Venice. It is of historical interest only.
Strengths & trade-offs
Strengths
- +Read-only stores could bulk-load precomputed Hadoop output efficiently
- +Simple Dynamo-style design that was easy to reason about
Trade-offs
- −Unmaintained since 2017 with no security fixes
- −Its original owner has migrated away
- −No modern features such as secondary indexes or TTL
Great for
- +Understanding Dynamo-style system design
- +Legacy systems awaiting migration
Think twice for
- −Any new project
Capabilities
Native or official first-party support · 1 of 25 tracked
Scale & consistency
- Comfortable data size
- Large · ≤ 1 PB
- Horizontal scaling
- Native
- Multi-region
- No
- Transactions
- None
- Consistency
- Eventual
- Schema
- Schemaless
- Architecture
- Distributed, Multi primary, Client server
- Query languages
- Voldemort client API
Deployment
Ecosystem
Alternatives to Project Voldemort
Erlang-based, Dynamo-style key-value store built to stay up through ugly failures.
project-voldemort vs riak-kv →Masterless wide-column database built for always-on writes across data centers.
project-voldemort vs apache-cassandra →C++ rewrite of Cassandra with shard-per-core design; also speaks the DynamoDB API.
project-voldemort vs scylladb →AWS's serverless key-value and document store with single-digit-millisecond reads at any scale.
project-voldemort vs amazon-dynamodb →Apple's ordered, strictly serializable distributed key-value store, built to have layers on top.
project-voldemort vs foundationdb →CNCF-graduated, Raft-replicated transactional key-value store in Rust; the storage layer of TiDB.
project-voldemort vs tikv →Project Voldemort FAQ
What is Project Voldemort?
Project Voldemort is a key-value database by LinkedIn, first released in 2009. Voldemort was LinkedIn's distributed key-value store, one of the first open-source implementations of Amazon's Dynamo paper, with consistent hashing, vector-clock versioning, pluggable storage engines and a well-known read-only mode for bulk-loading Hadoop output. It served LinkedIn features such as 'People You May Know' for years. LinkedIn stopped all production use in 2018 and the README states the project is no longer under development; the read-only use cases moved to Venice. It is of historical interest only.
Is Project Voldemort open source?
Yes. Project Voldemort is open source. License: Apache License 2.0.
Does Project Voldemort support ACID transactions?
No. Project Voldemort does not provide ACID transactions; its consistency model is eventual.
Does Project Voldemort scale horizontally?
Project Voldemort scales horizontally out of the box, with built-in sharding or partitioning. It is comfortable at large scale (≤ 1 PB).
How can I run Project Voldemort?
You can self-host it.
What is Project Voldemort best for?
Project Voldemort is a strong fit for understanding Dynamo-style system design and legacy systems awaiting migration. Think twice for any new project.
What are the best alternatives to Project Voldemort?
Common alternatives to Project Voldemort include Riak KV, Apache Cassandra, ScyllaDB, Amazon DynamoDB and FoundationDB. The right choice depends on your workload, scale and hosting; WhichDB's smart match ranks them against your requirements.
Is Project Voldemort still maintained?
Project Voldemort is discontinued. No longer developed; LinkedIn ended production use in 2018 and migrated to Venice. Last release 1.10.26 (2017).
How popular is Project Voldemort?
Project Voldemort ranks #348 of 844 databases on the WhichDB Score (7.1/100), and #49 among key-value databases. The score combines public adoption signals such as package downloads, GitHub activity, job posts and developer discussion.