Skip to content
Go to file

Latest commit


Git stats


Failed to load latest commit information.



LevelDB JNI gives you a Java interface to the LevelDB C++ library which is a fast key-value storage library written at Google that provides an ordered mapping from string keys to string values..

Getting the JAR

Just add the following jar to your java project: leveldbjni-all-1.8.jar

Using as a Maven Dependency

You just need to add the following dependency to your Maven POM:


By using the leveldbjni-all dependency, you get the OS specific native drivers for all supported platforms.

If you want to use only one or some but not all native drivers, then directly use the OS specific dependency instead of leveldbjni-all. For example to use Linux 64 bit, use this dependency:


If you have the leveljni native driver DLL/SO library already separately installed e.g. by a package manager (see issue 90), then you could depend on the Java "launcher" without the JAR containing the OS specific native driver like this:


Lastly, another project unrelated to this project separately provides a (less mature) pure Java implementation of LevelDB, see dain/leveldb. Note that both that and this project share the same Maven artefact for the Level DB API interface (org.iq80.leveldb:leveldb-api).

API Usage:

Recommended Package imports:

import org.iq80.leveldb.*;
import static org.fusesource.leveldbjni.JniDBFactory.*;

Opening and closing the database.

Options options = new Options();
DB db = File("example"), options);
try {
  // Use the db in here....
} finally {
  // Make sure you close the db to shutdown the 
  // database and avoid resource leaks.

Putting, Getting, and Deleting key/values.

db.put(bytes("Tampa"), bytes("rocks"));
String value = asString(db.get(bytes("Tampa")));

Performing Batch/Bulk/Atomic Updates.

WriteBatch batch = db.createWriteBatch();
try {
  batch.put(bytes("Tampa"), bytes("green"));
  batch.put(bytes("London"), bytes("red"));

} finally {
  // Make sure you close the batch to avoid resource leaks.

Iterating key/values.

DBIterator iterator = db.iterator();
try {
  for(iterator.seekToFirst(); iterator.hasNext(); {
    String key = asString(iterator.peekNext().getKey());
    String value = asString(iterator.peekNext().getValue());
    System.out.println(key+" = "+value);
} finally {
  // Make sure you close the iterator to avoid resource leaks.

Working against a Snapshot view of the Database.

ReadOptions ro = new ReadOptions();
try {
  // All read operations will now use the same 
  // consistent view of the data.
  ... = db.iterator(ro);
  ... = db.get(bytes("Tampa"), ro);

} finally {
  // Make sure you close the snapshot to avoid resource leaks.

Using a custom Comparator.

DBComparator comparator = new DBComparator(){
    public int compare(byte[] key1, byte[] key2) {
        return new String(key1).compareTo(new String(key2));
    public String name() {
        return "simple";
    public byte[] findShortestSeparator(byte[] start, byte[] limit) {
        return start;
    public byte[] findShortSuccessor(byte[] key) {
        return key;
Options options = new Options();
DB db = File("example"), options);

Disabling Compression

Options options = new Options();
DB db = File("example"), options);

Configuring the Cache

Options options = new Options();
options.cacheSize(100 * 1048576); // 100MB cache
DB db = File("example"), options);

Getting approximate sizes.

long[] sizes = db.getApproximateSizes(new Range(bytes("a"), bytes("k")), new Range(bytes("k"), bytes("z")));
System.out.println("Size: "+sizes[0]+", "+sizes[1]);

Getting database status.

String stats = db.getProperty("leveldb.stats");

Getting informational log messages.

Logger logger = new Logger() {
  public void log(String message) {
Options options = new Options();
DB db = File("example"), options);

Destroying a database.

Options options = new Options();
factory.destroy(new File("example"), options);

Repairing a database.

Options options = new Options(); File("example"), options);

Using a memory pool to make native memory allocations more efficient:

JniDBFactory.pushMemoryPool(1024 * 512);
try {
    // .. work with the DB in here, 
} finally {


See also


Supported Platforms

The following worked for me on:

  • OS X Lion with X Code 4
  • CentOS 5.6 (32 and 64 bit)
  • Ubuntu 12.04 (32 and 64 bit)
  • apt-get install autoconf libtool

Build Procedure

Then download the snappy, leveldb, and leveldbjni project source code:

tar -zxvf snappy-1.0.5.tar.gz
git clone git://
git clone git://
export SNAPPY_HOME=`cd snappy-1.0.5; pwd`
export LEVELDB_HOME=`cd leveldb; pwd`
export LEVELDBJNI_HOME=`cd leveldbjni; pwd`

Compile the snappy project. This produces a static library.

./configure --disable-shared --with-pic

Patch and Compile the leveldb project. This produces a static library.

git apply ../leveldbjni/leveldb.patch
make libleveldb.a

Now use maven to build the leveldbjni project.

mvn clean install -P download -P ${platform}

Replace ${platform} with one of the following platform identifiers (depending on the platform your building on):

  • osx
  • linux32
  • linux64
  • win32
  • win64
  • freebsd64

If your platform does not have the right auto-tools levels available just copy the leveldbjni-${version} artifact from a platform the does have the tools available then add the following argument to your maven build:


Build Results

  • leveldbjni/target/leveldbjni-${version}.jar : The java class file to the library.
  • leveldbjni/target/leveldbjni-${version} : A GNU style source project which you can use to build the native library on other systems.
  • leveldbjni-${platform}/target/leveldbjni-${platform}-${version}.jar : A jar file containing the built native library using your currently platform.
You can’t perform that action at this time.