Skip to content
An open source clone of Amazon's Dynamo.
Java C++ Python Shell Ruby Protocol Buffer Other
Find file
Failed to load latest commit information.
.settings Format only edited lines setting
bin Removed trailing whitespace, made list of partitions sorted by intege…
clients fix
config Minimal config for read-only store, with Hadoop (BnP) hints.
contrib Better exception handling in HdfsFetcher#fetchFromSource().
docs Updating omnigraffle for logical architecture
example Polish up the Java Client example
gradle/wrapper gradle build fails on mac
private-lib Using rocksdbjni jar from maven
src Bug fixes and improvements for Build and Push.
test ClientRequestExecutor Pool test unit test failure
voldemort-contrib Dummy directory for injecting custom gradle behavior
voldemort-protobuf Gradle Protobuf shadowed jar
.gitignore Migrate from stores.xml to STORES folder in all example configs
LICENSE Add license header to source files.
NOTES Edited NOTES to correct the locate command (which is now preflist).
NOTICE Add Avro, Jackson and ParaNamer to NOTICE. Fix the Readme to use Gradle
build.gradle Gradle Protobuf shadowed jar
build.xml Fully disabled the Ant build in favor of the Gradle one. Releasing Voldemort 1.10.13
gradlew Generate gradle wrapper to work with versions
gradlew.bat Generate gradle wrapper to work with versions
release_notes.txt Releasing Voldemort 1.10.13
settings.gradle Added empty settings.gradle so that Gradle does not search for it in … Initial import
web.xml Initial import

Voldemort is a distributed key-value storage system


  • Data is automatically replicated over multiple servers across multiple datacenters.
  • Data is automatically partitioned so each server contains only a subset of the total data
  • Server failure is handled transparently
  • Pluggable serialization is supported to allow rich keys and values including lists and tuples with named fields, as well as to integrate with common serialization frameworks like Protocol Buffers, Thrift, and Java Serialization
  • Data items are versioned to maximize data integrity in failure scenarios without compromising availability of the system
  • Each node is independent of other nodes with no central point of failure or coordination
  • Pluggable storage engines, to cater to different workloads
  • SSD Optimized Read Write storage engine, with support for multi-tenancy
  • Built in mechanism to fetch & serve batch computed data from Hadoop
  • Support for pluggable data placement strategies to support things like distribution across data centers that are geographical far apart.

It is used at LinkedIn by numerous critical services powering a large portion of the site. .


You can refer to for more info

Download Code

cd ~/workspace
git clone
cd voldemort
./gradlew clean jar

Start Server

# in one terminal
bin/ config/single_node_cluster

Use Client Shell

Client shell gives you fast access to the store. We already have a test store defined in the "single_node_cluster", whose key and value are both String.

# in another terminal
cd ~/workspace/voldemort
bin/ test tcp://localhost:6666/

Now you have the the voldemort shell running. You can try these commands in the shell

put "k1" "v1"
put "k2" "v2"
get "k1"
getall "k1" "k2"
delete "k1"
get "k1"

You can find more commands by runninghelp

Comparison to relational databases

Voldemort is not a relational database, it does not attempt to satisfy arbitrary relations while satisfying ACID properties. Nor is it an object database that attempts to transparently map object reference graphs. Nor does it introduce a new abstraction such as document-orientation. It is basically just a big, distributed, persistent, fault-tolerant hash table. For applications that can use an O/R mapper like ActiveRecord or Hibernate this will provide horizontal scalability and much higher availability but at great loss of convenience. For large applications under internet-type scalability pressure, a system may likely consist of a number of functionally partitioned services or apis, which may manage storage resources across multiple data centers using storage systems which may themselves be horizontally partitioned. For applications in this space, arbitrary in-database joins are already impossible since all the data is not available in any single database. A typical pattern is to introduce a caching layer which will require hashtable semantics anyway. For these applications Voldemort offers a number of advantages:

  • Voldemort combines in memory caching with the storage system so that a separate caching tier is not required (instead the storage system itself is just fast).
  • Unlike MySQL replication, both reads and writes scale horizontally
  • Data partioning is transparent, and allows for cluster expansion without rebalancing all data
  • Data replication and placement is decided by a simple API to be able to accommodate a wide range of application specific strategies
  • The storage layer is completely mockable so development and unit testing can be done against a throw-away in-memory storage system without needing a real cluster (or even a real storage system) for simple testing


The source code is available under the Apache 2.0 license. We are actively looking for contributors so if you have ideas, code, bug reports, or fixes you would like to contribute please do so.

For help please see the discussion group, or the IRC channel #voldemort. Bugs and feature requests can be filed on Github.

Special Thanks

We would like to thank JetBrains for supporting Voldemort Project by offering open-source license of their IntelliJ IDE to us.

Something went wrong with that request. Please try again.