Java IoT Authors: Elizabeth White, Liz McMillan, Yeshim Deniz, Pat Romanski, Roger Strukhoff

Related Topics: Java IoT

Java IoT: Article

Building Applications with Berkeley DB Java Edition

High-performance database goes pure Java

Berkeley DB is a database with a long history. First released in 1991 as a replacement for various dbm implementations, it was soon included in BSD Unix releases. Requests for new features and commercial support led to the formation of Sleepycat Software in 1996. Using a dual license model, Berkeley DB became very popular for both open source and commercial applications. Recently, Sleepycat announced Berkeley DB Java Edition (JE), a 100% Java implementation that runs on any J2SE 1.4.2 or later JVM.

This article introduces JE's features and classes and uses a sample electronic voting application to illustrate its use.

Berkeley DB and JE
The original Berkeley DB is used in a wide variety of open source and commercial applications in telecom, Internet infrastructure, storage, security, financial services, and other industries. Sleepycat estimates that there are over 200 million copies deployed, but most people are not even aware they touched Berkeley DB. (See "Where's the Sleeping Cat?" available at the URL listed in References.)

Berkeley DB is a nonrelational database that stores key/value pairs. Databases can be configured to store multiple values under a single key. Secondary databases can be created that use different keys to access the same values. Searches and joins can be performed using a specific key or a range of keys.

The original Berkeley DB (not JE) is written in C, though there are bindings for other languages, including Java. JE brings to Java developers the features of Berkeley DB and adds better integration with the Java world through Java data type support while avoiding the performance penalty of JNI.

Like its C ancestor, JE is small. The database's footprint is less than 435K. It supports full ACID transactions (meaning its characteristics include Atomicity, Consistency, Isolation, and Durability), performs record-level locking for high concurrency, and can handle huge amounts of data: hundreds of terabytes in a single table, with record sizes up to two gigabytes. Frequently used data is cached in memory. JE has better concurrency than the C Berkeley DB for some applications because, unlike DB, it has record-level locking that allows applications to continue accessing other parts of the database when a record is locked.

JE databases are stored as files on the file system. Backing up a database is accomplished by copying the files. To restore a database, move the copied files into the database directory and restart the application. When an application uses a database with transactions (that group database operations together in order to treat them atomically), checkpoints are periodically saved to the file system by a separate JE thread in order to make recovery faster. Applications without transactions may need to perform manual syncs that tell the database to flush data in memory to the file system, depending upon their requirements.

Technology Overview
JE databases are B+Trees whose records are key/value pairs. The keys and values are byte arrays. Databases are stored as files within a single directory. An in-memory cache stores as much as it can of the B+Tree structure and frequently accessed data.

The class DatabaseStats provides information about a database, such as the number of records in the database or the depth of the B+Tree.

A database environment coordinates one or more databases. It manages transactions, allowing them to work across databases. The environment provides the in-memory cache used by an application. It also facilitates administrative operations such as renaming or deleting databases.

Records and Binding
Each record in a database is an instance of DatabaseEntry, essentially a wrapper around a byte array. Since storing Java objects is a common operation, JE provides ways to translate between them and byte arrays. This process is called binding. JE provides classes for binding simple numeric and string objects.

There are two approaches to creating DatabaseEntry values from more complex Java objects: serialization or creating a custom class that translates the objects to and from byte arrays. When serializing objects to be stored in separate database records, each serialized instance carries with it duplicate information: the descriptions of the classes involved in the serialization. JE provides binding APIs that perform the serialization and store the duplicate information only once, in a separate database (see Figure 1).

A custom binding that maps objects to byte arrays must subclass TupleBinding, which implements the EntryBinding interface. This interface consists of two methods: objectToEntry and entryToObject.

Cursors and Iterators
To look up a single key, call myDatabase.get(), passing in the key as a DatabaseEntry and another DatabaseEntry to hold the returned data.

To iterate through the database or a subset of the database, both a cursor and iterators are available. They operate the same way, searching for a (possibly partial) key and iterating over the returned values. The difference is that iteration uses flavors of the Java collection classes called StoredMap, StoredSet, and StoredIterator. Unlike java.util.Iterators, StoredIterators must be closed explicitly. There are two ways to close a StoredIterator: call its close method or call the static method StoredIterator.close. The latter approach avoids casting when the iterator is stored in a variable of type Iterator.

The JE collection classes implement all of the collection methods except for size (for example, get, put, and containsValue) and a few extras. See the Javadocs for details.

Secondary Databases
To perform searches using keys that are different than those used to store the data, a secondary database must be created. The values in the secondary database are the same as those in the primary database. Multiple secondary databases may be created.

When a primary database is modified, JE updates all of its secondary databases. All writing takes place through the primary database. The one exception to this rule is that records may be deleted from a secondary database. The records in the primary database are also deleted, as are all corresponding records in all other secondary databases.

Creating a secondary database requires three things: a primary database, a binding for the keys, and a binding for the values. The value binding will be the same one used by the primary database. The key creator it must implement is the SecondaryKeyCreator interface, which has the single method createSecondaryKey.

Transactions are enabled by an application when an environment and its databases are created or opened. Once enabled, they must be used for all database modifications. Methods such as Database.get and Database.put take a transaction object as an argument. If the transaction argument is null and transactions are enabled, the method may either use autocommit (committing the data when the operation is finished) as does "put" or it may not use a transaction at all, as with "get", which does not modify the database.

The TransactionRunner class and TransactionWorker interface can make using transactions easier by handling retries and exceptions. A TransactionRunner creates a transaction and calls TransactionWorker.doWork. The runner can retry the work any number of times.

A Sample Application
A pair of electronic voting applications - one that simulates election day and another that runs reports against the collected data - will help illustrate the use of JE. While developing this application, I made heavy use of Sleepycat's JE Javadocs and "Getting Started with Berkeley DB Java Edition," included in the documentation that comes with JE.

The two applications share much of their code. To create the data, the election day simulator creates voting booths, then generates votes and sends them to the vote server. The vote server stores information about the booths and their votes into multiple primary and secondary databases. The reporting application performs queries about the booths and the votes.

The three main classes for these applications are Booth, Vote, and VoteServer. A booth is identified by an IP address and knows what state, city, and district it is in. The booth assigns each vote a sequence number. A vote's unique identifier is a combination of the booth's IP address and the vote's sequence number. A vote holds a few more fields like the race (President, State Assembly, Dog Catcher), political party (Democrat, Republican, Silly), and candidate.

After thinking about the queries I wanted to be able to run, I decided that in addition to the primary databases for Booth and Vote, I would need three secondary databases: for booths using state as the key, for votes using race as the key, and for votes using race plus political party.

Before creating databases, bindings for Booth and Vote need to be created. For the primary databases, I chose to use custom binding classes. Listing 1 contains the two methods implemented by BoothBinding. The JE classes TupleInput and TupleOutput know how to read and write strings and intrinsics (ints, longs, etc.).

Bindings for the keys in the primary databases are unnecessary because the keys are so simple: the booth database's key is the IP address byte array that needs no translation, and the vote database's key is a string from which we can retrieve a byte array.

Each secondary database needs a SecondaryKeyCreator. One of the vote secondary databases is indexed by race and party. Listing 2 shows the method that, given a vote and its primary key, fills in the secondary key DatabaseEntry.

When opening a database for writing, it is created if it does not exist. A flag is set that determines if the database is transactional. Listing 3 shows this process.

Storing Votes
The code in Listing 4 creates a transaction, calls the method that stores the booth and its votes, and commits the transaction. (Listings 4-9 can be downloaded from www.sys-con.com/java/sourcec.cfm.) This code commits without synchronizing the change to the log file, which is risky (the data is held in memory until the next sync) but faster.

Booth and vote data are written to the primary database. The secondary databases are updated automatically. The code in Listing 5 stores a vote into the vote primary database.

The VoteServer runs five different reports, each using a different technique.

Total Votes
The following code retrieves the total number of votes in the primary vote database.

// try/catch not shown
Database db = dbEnvironment.getVoteDb();
Cursor cursor =
  db.openCursor(null, null);

DatabaseEntry foundKey =
  new DatabaseEntry();
DatabaseEntry foundData =
  new DatabaseEntry();
int numRecords = 0;
while (cursor.getNext(foundKey,
    == OperationStatus.SUCCESS)
int nRecs = dbEnvironment.getVoteDb()
// (print number of records here)

Single Vote
Database.get is used to find a single vote by its primary key (see Listing 6).

Booths by State
This report selects a state from the database, then prints all of the booths in that state. It runs the same query two different ways: with a cursor (see Listing 7) and with a JE Iterator (see Listing 8). The Iterator code is shorter. In both listings, error handling code and try/catch blocks are not shown.

Total Presidential Race Vote Count
To count how many votes were cast in the presidential race, a StoredMap is opened on the proper secondary database. The duplicate records (votes) for the presidential race are retrieved and the number of votes is reported (see Listing 9).

Presidential Race Results
The last report prints the total votes per party for the presidential race and declares a winner. As in the previous report, a map is created on a secondary database and the map's size method returns the vote count.

The electronic voting application source code contains plenty of comments and deals with all the details ignored in this article such as error handling. The code and properties files that configure the databases and transactions have not been tuned for performance.

To get started with the Berkeley DB Java Edition, download it from Sleepycat Software. Read "Getting Started with Berkeley DB Java Edition," browse the example code that comes with JE, and refer to the JE Javadocs.

JE is available under a dual license. The open source license is online at www.sleepycat.com/download/oslicense.html. Pricing starts at $40,000 for a buyout license, which enables the customer to redistribute JE within a specific product/service in any volume into perpetuity.

Sleepycat Software's site is www.sleepycat.com. There you can download the Berkeley DB Java Edition. The documentation and example code that comes with JE are great starting points. The site also contains a number of technical articles and white papers at www.sleepycat.com/company/technical.shtml.

The paper "Where's the Sleeping Cat?" is available at www.sleepycat.com/pdfs/index.php?paper=jdj_wheresthecat.

The latest version of the electronic voting application code can be found at www.io.com/~jimm/writing/evote.tar.gz.


While creating or opening databases, JE allows you to specify many optional database configuration parameters, including but not limited to:

  • Maximum database file size
  • Whether transactions are allowed
  • Transaction timeout length
  • Cache size
  • Whether duplications are allowed
  • How frequently to perform checkpoints
  • How to sort duplicate records
All of this is done through the Java API. You don't need a database administrator to create or maintain your databases.


Why Berkeley DB Java Edition?
Where should you use Berkeley DB Java Edition? Consider using JE if:

  • Speed and size are issues. JE is fast and can handle huge amounts of data.
  • It makes sense to embed the database into your application. Some applications are delivered to platforms that don't have a database installed or can't connect to a database server.
  • You know what questions you will be asking of the data. Writing queries involves writing Java code and designing the database structure to allow access to the data based on keys other than the primary key.
JE may not be the solution if:
  • The application needs to perform ad-hoc queries. For this you need to write Java code to perform searches; there is no other query language. Searches that use keys other than the primary key used to store the data require a secondary database, discussed later in this article.
  • Data must be shared with non-Java applications. The original Berkeley DB data format is different than the JE format. Berkeley DB supports bindings to many languages, including Java.

Database Log Files
Databases are stored as a series of log files. When data is written to a disk, it's always appended to an existing log file. This enables very fast writes since the disk head does not need to move. Log files are never overwritten or modified; those operations are slower than appends. When the log file in use by a database reaches a maximum size, a new log file is created. A background "cleaner" thread organizes the log files, moving active records from older log files to newer ones and deleting old log files that contain no active data.

More Stories By Jim Menard

Jim Menard is a senior technologist with over 20 years of experience in software development, design, and management. He has developed many open source projects including the Java GUI report writer DataVision (http://datavision.sourceforge.net). Jim likes shiny things (http://www.io.com/~jimm).

Comments (3) View Comments

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.

Most Recent Comments
Jim Menard 10/06/04 04:31:59 PM EDT


I haven't tried creating a hierarchical structure using Berkeley DB. I imagine you could by containing the key of one database within the value of another.

I highly recommend the Sleepycat email lists. The "bdbje" email list is read by users and Sleepycat people (Sleepycats?) alike. See http://www.sleepycat.com/supports/index.shtml for the list of mailing lists. To join the bdbje list, send an email to [email protected].

Lars Borup Jensen 10/06/04 04:16:17 AM EDT

Hi Jim,

Ive been looking for articles regarding berkeley db, IMHO a underused product. I've been wondering for a looong time now, how it would be possible to create a hierarchical structure using bdb. Ive looked at openldap (their bdb and hdb backends) but still "designing" as this task does NOT seem trivial. Have you had any experience regarding this??

Jim Menard 09/08/04 02:02:26 PM EDT

Please ignore the text between "[DELETE" and "END DELETE]" in the sample code here online. It's the artifact of a last-minute change. I haven't yet seen the article in its print form, so I'm not sure if the printed code has the same issue.

I look forward to your comments on this article. Thanks for reading it.

@ThingsExpo Stories
Cloud-enabled transformation has evolved from cost saving measure to business innovation strategy -- one that combines the cloud with cognitive capabilities to drive market disruption. Learn how you can achieve the insight and agility you need to gain a competitive advantage. Industry-acclaimed CTO and cloud expert, Shankar Kalyana presents. Only the most exceptional IBMers are appointed with the rare distinction of IBM Fellow, the highest technical honor in the company. Shankar has also receive...
Enterprises have taken advantage of IoT to achieve important revenue and cost advantages. What is less apparent is how incumbent enterprises operating at scale have, following success with IoT, built analytic, operations management and software development capabilities - ranging from autonomous vehicles to manageable robotics installations. They have embraced these capabilities as if they were Silicon Valley startups.
Poor data quality and analytics drive down business value. In fact, Gartner estimated that the average financial impact of poor data quality on organizations is $9.7 million per year. But bad data is much more than a cost center. By eroding trust in information, analytics and the business decisions based on these, it is a serious impediment to digital transformation.
The standardization of container runtimes and images has sparked the creation of an almost overwhelming number of new open source projects that build on and otherwise work with these specifications. Of course, there's Kubernetes, which orchestrates and manages collections of containers. It was one of the first and best-known examples of projects that make containers truly useful for production use. However, more recently, the container ecosystem has truly exploded. A service mesh like Istio addr...
Predicting the future has never been more challenging - not because of the lack of data but because of the flood of ungoverned and risk laden information. Microsoft states that 2.5 exabytes of data are created every day. Expectations and reliance on data are being pushed to the limits, as demands around hybrid options continue to grow.
Business professionals no longer wonder if they'll migrate to the cloud; it's now a matter of when. The cloud environment has proved to be a major force in transitioning to an agile business model that enables quick decisions and fast implementation that solidify customer relationships. And when the cloud is combined with the power of cognitive computing, it drives innovation and transformation that achieves astounding competitive advantage.
As IoT continues to increase momentum, so does the associated risk. Secure Device Lifecycle Management (DLM) is ranked as one of the most important technology areas of IoT. Driving this trend is the realization that secure support for IoT devices provides companies the ability to deliver high-quality, reliable, secure offerings faster, create new revenue streams, and reduce support costs, all while building a competitive advantage in their markets. In this session, we will use customer use cases...
Digital Transformation: Preparing Cloud & IoT Security for the Age of Artificial Intelligence. As automation and artificial intelligence (AI) power solution development and delivery, many businesses need to build backend cloud capabilities. Well-poised organizations, marketing smart devices with AI and BlockChain capabilities prepare to refine compliance and regulatory capabilities in 2018. Volumes of health, financial, technical and privacy data, along with tightening compliance requirements by...
The best way to leverage your Cloud Expo presence as a sponsor and exhibitor is to plan your news announcements around our events. The press covering Cloud Expo and @ThingsExpo will have access to these releases and will amplify your news announcements. More than two dozen Cloud companies either set deals at our shows or have announced their mergers and acquisitions at Cloud Expo. Product announcements during our show provide your company with the most reach through our targeted audiences.
DevOpsSummit New York 2018, colocated with CloudEXPO | DXWorldEXPO New York 2018 will be held November 11-13, 2018, in New York City. Digital Transformation (DX) is a major focus with the introduction of DXWorldEXPO within the program. Successful transformation requires a laser focus on being data-driven and on using all the tools available that enable transformation if they plan to survive over the long term. A total of 88% of Fortune 500 companies from a generation ago are now out of bus...
With 10 simultaneous tracks, keynotes, general sessions and targeted breakout classes, @CloudEXPO and DXWorldEXPO are two of the most important technology events of the year. Since its launch over eight years ago, @CloudEXPO and DXWorldEXPO have presented a rock star faculty as well as showcased hundreds of sponsors and exhibitors! In this blog post, we provide 7 tips on how, as part of our world-class faculty, you can deliver one of the most popular sessions at our events. But before reading...
DXWordEXPO New York 2018, colocated with CloudEXPO New York 2018 will be held November 11-13, 2018, in New York City and will bring together Cloud Computing, FinTech and Blockchain, Digital Transformation, Big Data, Internet of Things, DevOps, AI, Machine Learning and WebRTC to one location.
DXWorldEXPO LLC announced today that ICOHOLDER named "Media Sponsor" of Miami Blockchain Event by FinTechEXPO. ICOHOLDER give you detailed information and help the community to invest in the trusty projects. Miami Blockchain Event by FinTechEXPO has opened its Call for Papers. The two-day event will present 20 top Blockchain experts. All speaking inquiries which covers the following information can be submitted by email to [email protected] Miami Blockchain Event by FinTechEXPO also offers s...
DXWorldEXPO | CloudEXPO are the world's most influential, independent events where Cloud Computing was coined and where technology buyers and vendors meet to experience and discuss the big picture of Digital Transformation and all of the strategies, tactics, and tools they need to realize their goals. Sponsors of DXWorldEXPO | CloudEXPO benefit from unmatched branding, profile building and lead generation opportunities.
Dion Hinchcliffe is an internationally recognized digital expert, bestselling book author, frequent keynote speaker, analyst, futurist, and transformation expert based in Washington, DC. He is currently Chief Strategy Officer at the industry-leading digital strategy and online community solutions firm, 7Summits.
Digital Transformation and Disruption, Amazon Style - What You Can Learn. Chris Kocher is a co-founder of Grey Heron, a management and strategic marketing consulting firm. He has 25+ years in both strategic and hands-on operating experience helping executives and investors build revenues and shareholder value. He has consulted with over 130 companies on innovating with new business models, product strategies and monetization. Chris has held management positions at HP and Symantec in addition to ...
The IoT Will Grow: In what might be the most obvious prediction of the decade, the IoT will continue to expand next year, with more and more devices coming online every single day. What isn’t so obvious about this prediction: where that growth will occur. The retail, healthcare, and industrial/supply chain industries will likely see the greatest growth. Forrester Research has predicted the IoT will become “the backbone” of customer value as it continues to grow. It is no surprise that retail is ...
Andrew Keys is Co-Founder of ConsenSys Enterprise. He comes to ConsenSys Enterprise with capital markets, technology and entrepreneurial experience. Previously, he worked for UBS investment bank in equities analysis. Later, he was responsible for the creation and distribution of life settlement products to hedge funds and investment banks. After, he co-founded a revenue cycle management company where he learned about Bitcoin and eventually Ethereal. Andrew's role at ConsenSys Enterprise is a mul...
DXWorldEXPO LLC announced today that "Miami Blockchain Event by FinTechEXPO" has announced that its Call for Papers is now open. The two-day event will present 20 top Blockchain experts. All speaking inquiries which covers the following information can be submitted by email to [email protected] Financial enterprises in New York City, London, Singapore, and other world financial capitals are embracing a new generation of smart, automated FinTech that eliminates many cumbersome, slow, and expe...
Cloud Expo | DXWorld Expo have announced the conference tracks for Cloud Expo 2018. Cloud Expo will be held June 5-7, 2018, at the Javits Center in New York City, and November 6-8, 2018, at the Santa Clara Convention Center, Santa Clara, CA. Digital Transformation (DX) is a major focus with the introduction of DX Expo within the program. Successful transformation requires a laser focus on being data-driven and on using all the tools available that enable transformation if they plan to survive ov...