There are many open source reporting tools readily available for companies to deploy to their business users today.  Most of them, however, do not provide hard core corporate features such as multiple user support, integration with existing security mechanisms, ad hoc visual reporting tools, scheduling facilities, and the like.  This comparison will briefly examine some of the more common contenders before focusing on the primary candidates for enterprise deployment.

Tools addressed in this comparison include:

• BIRT
• DataVision
• Jacob
• JasperReports
• Jedox
• JFreeReport
• jXLS
• OpenReport
• OpenReports
• Pentaho
• POI

Background

Open source reporting tools have been around for a number of years.  They provide functionality ranging from developer-only libraries that help organize relational data to visual report designers for IT professionals to web-based ad hoc querying applications targeted at business users.  Only in the past two years have some of these packages matured to a point where an enterprise can reasonably select one as a significant piece of technology to be deployed throughout the organization.

Although developer libraries and visual report designers from the open source world can be very powerful, it is the web-based, business user-facing applications that make this category interesting from an enterprise perspective.  The key here is that business users can frequently serve themselves by choosing, running, viewing, and saving reports without help from IT.  These abilities imply a set of criteria by which we can compare the open source reporting tools:

• Ease of use:  the need for training should be minimal and the interface should be pleasurable to use
• Ease of deployment:  business users should need only a standard web browser to access the tool
• Security:  the application should work with established security mechanisms (e.g., LDAP)
• Distribution:  it should be easy for users to have reports emailed to themselves and to others
• Scheduling:  end users should be able to schedule future report runs (e.g., first Monday of each month)
• Multiple formats:  users like to have options when exporting report data (e.g., XLS, PDF, CSV)
• Availability:  users want the reporting system available 24×7 (i.e., clusterability, scalability)
• Ease of administration:  it should be straightforward to set up and maintain the solution
• Viability:  the tool should have a well-structured and thriving community with a reasonable license

Contenders

Before focusing on the primary candidates for enterprise roll out, it is important to cover other reporting tools that are frequently used by developers, because not all reporting needs are of enterprise scope.

DataVision and OpenReport

DataVision and OpenReport are examples of developer libraries that aid developers in building reporting functionality into their applications.  As such, they are necessarily limited in addressing the criteria we set out above as important for enterprise business users.

jXLS, POI, and Jacob

These tools and others like them allow Java developers to create and/or manipulate XLS files to help with Microsoft Excel integration.  They are fairly low level and do not meet our enterprise criteria.

Jedox

Available as an open source component, Jedox provides “enterprise spreadsheets” so that end users can run reports inside of Microsoft Excel.  This works through an Excel plug-in and does not provide a separate web-based interface.  As such, it does not meet the “ease of deployment” criterion.

JFreeReport

JFreeReport is now part of the Pentaho project and so it will be covered in the context of Pentaho later in this comparison.

BIRT

BIRT, Business Intelligence Reporting Tools, is a sub-project of Eclipse.  It was originally donated by Actuate, a commercial provider of primarily proprietary reporting tools.  BIRT focuses on a report designer that is available as an Eclipse plug-in.  It also includes a debugger and other development tools.  Enterprise deployment of BIRT-based reports as defined by our criteria above is only available at this time through the proprietary iServer product from Actuate and is therefore not addressed further in this comparison of open source reporting tools for the enterprise.

Primary Candidates

Three open source projects stand out as potential enterprise reporting solutions because they address each of the criteria set out at the beginning of this comparison.  The candidates are:

• JasperReports
• OpenReports
• Pentaho

OpenReports

Not to be confused with OpenReport, OpenReports (oreports.com) provides enterprise wrappings around a number of reporting engines that include:  BIRT, JasperReports, JFreeReport/Pentaho, and jXLS.  It offers a web-based interface with support for scheduling, security integration, and administration.  The primary benefit of OpenReports is that it allows users to mix and match reporting engines.  In this fashion, different groups within an enterprise are able to use different report design tools yet share the same server platform.

OpenReports is licensed under the GPL, version 2.  It is based on enterprise Java technology, supports OLAP functionality, and offers many report formats that include HTML, PDF, CSV, XLS, RTF, and image.  It allows reports to be scheduled and sent to users via email.  It also has both SOAP- and REST-based API’s.

The Professional version of OpenReports adds the following features:  user dashboards that can display multiple reports, charts, or graphs simultaneously; drilldown charts; and report usage statistics.  The list price for this version is approximately $500 per server.

Unfortunately, OpenReports is being developed by a single programmer and therefore does not meet our criterion for “viability”.

Pentaho

One of the two primary enterprise-ready open source reporting tools with significant commercial backing, Pentaho focuses on data integration and workflow automation.  By the way, the name “Pentaho” was created when the company was formed in 2004 by five original founders who wanted to retire in Lake Tahoe.

Overview

The Pentaho Business Intelligence suite includes:

• Reporting (Report designer and viewer)
• Analysis (OLAP)
• Dashboards (Portal, metrics, alerts)
• Data Mining (Relationship and trend discovery)

Pentaho is composed of a number of open source components, which has led to a number of open source licenses.  Its various top level components are licensed under either the GPL version 2, the LGPL, or the MPL (Mozilla Public License).  If you intend to redistribute Pentaho as part of your commercial offering, pay close attention to the individual license(s) associated with each part you plan to deploy.

Technology

Pentaho has a Java Enterprise Edition (JEE) foundation that is based on JBoss.  In addition, it relies heavily on JBoss Portal for much, but not all, of its user interface.  This causes some undesirable seams in the user experience which will be addressed later.  To aid in security integration, it supports the standard Acegi open source package that allows pluggable authentication and authorization schemes.  It also integrates with CAS (Central Authentication Service) servers to facilitate single sign-on with other enterprise systems.

Features

Pentaho has what seems to be the “standard” set of open source reporting features, such as:

• Visual report designer and editor
• Web-based interface for business users to see, run, and export reports
• Several reporting formats, including HTML, PDF, XLS, and CSV
• Administrative functions, such as scheduling reports to run at certain times
• The ability to email reports to a number of users
• Users, groups, and role management to restrict report access to particular users
• Web service API’s for common functionality

Some of Pentaho’s strengths relative to its competition are:

• Web-based ad hoc reporting wizard:  lets technically-inclined business users point, click, drag, and drop their way to nice looking on-the-fly reports
• Conditional report distribution:  allows users to email reports when certain conditions are met (e.g., sales are below expectations by 25%)
• Task bar notification:  Windows users can get notified when a new report is ready to view
• Focus on data integration through ETL (Extract-Transform-Load) tools and automated workflows
• Data mining tools

Some noteworthy feature limitations:

• Administrators can only edit existing report schedules, not create new ones

The commercial version of the package, called Pentaho Professional, includes these additional features:

• Single Sign-On (SSO) support
• LDAP and Microsoft Active Directory (MSAD) integration
• Clustering support
• Report versioning
• Audit trail collection

Note:  The version of Pentaho tested was 1.7, released in May.  The current generally available production version as of the time of this writing is 1.7.1.

Usability

Web interface usability is the area in which Pentaho has the most room for improvement.  It is certainly powerful with report retrieval, parameter passing, drill-through charts, text linking, and administrative capabilities.  Unfortunately, it’s not always intuitive.  Due to the partial integration of JBoss Portal, it often feels like the user has to jump between two entirely different user interfaces.  That makes it difficult for first-time users to understand where to look for certain features, how to navigate through common screens, and how to get back to places they have already been.

The visual report designer and editor is good, but a little bit quirky.  As with its competition, it will take some time to get used to the occasional strange behavior and bake the workarounds into the standard workflow.

In general, Pentaho’s web interface is usable with some training and familiarization, but it leaves a lot to be desired.

Screenshots

Pentaho report designer

Pentaho web interface

Pentaho web report viewer

Jasper

Jasper reports was created as an open source project by Teodor Danciu in 2001.  Its feature set revolves around reporting and analysis with an overall emphasis on ease of use.

Overview

The Jasper Business Intelligence suite includes:

• JasperReports (Report designer and viewer)
• JasperServer (Admin, scheduler, ad hoc queries)
• JasperAnalysis (OLAP)
• JasperETL (Data integration)

Although Jasper is composed of a number of open source components, each top level component is licensed under the GPL version 2 or the LGPL.  If you intend to redistribute Jasper as part of your commercial offering, pay close attention to the terms of the license(s) that affect each component you plan to deploy.

Technology

Jasper has a Java foundation that is based on Tomcat, Spring, and Hibernate.  It supports the standard Acegi open source package to facilitate integration with existing authentication and authorization mechanisms.  As such, it can work in a single sign-on environment with other enterprise systems.

Features

Jasper has the “standard” set of open source reporting features, such as:

• Visual report designer and editor
• Web-based interface for business users to see, run, and export reports
• Several reporting formats, including HTML, PDF, XLS, and CSV
• Administrative functions, such as scheduling reports to run at certain times
• The ability to email reports to a number of users
• Users, groups, and role management to restrict report access to particular users
• Web service API’s for common functionality

Some of Jasper’s strengths relative to its competition are:

• Very good user interface
• Strong internationalization (I18N) support
• Hibernate query support in addition to SQL queries
• Custom report expressions can be written in Groovy, a Java-based scripting language
• Single sign-on support
• Fine-grained security model on OLAP reports

Some noteworthy feature limitations:

• Hibernate queries are powerful, but can lead to performance issues unless special care is paid to optimization (http://www.javalobby.org/articles/hibernatequery103/)

The commercial version of the package, called JasperReport Professional, includes these additional
features:

• End-user ad hoc querying and reporting
• Easy end-user created dashboards and mashups

Note:  The version of the Jasper suite tested was 3.0.0, released in June.  The current generally available
production version as of the time of this writing was 3.0.0.

Usability

Jasper’s web interface is not perfect, but it’s very good relative to its competition.  Most of it is a modern web 2.0 application with a clean and appealing look and feel.  It provides easy report retrieval, parameter passing, drill-through charts, text linking, and administrative capabilities.  It also allows users to export their reports in a multitude of formats, including XML and Flash in addition to the standard set (HTML, PDF, XLS, and CSV).  Jasper makes it easy for business users to schedule their own reports to run at certain times, such as the first Monday of each month.

For developers creating reports, the Jasper report designer can be used in either stand-alone mode or as a NetBeans plug-in.  As with other open source report designers, there are quirks that need to be worked around until they become habit.

In general, Jasper delivers a user experience far better than its competitors.

Screenshots

Jasper report designer

Jasper web admin console

Jasper report viewer

OLAP - Pentaho and Jasper

Both Pentaho and Jasper rely on Mondrian for OLAP (On-Line Analytical Processing) support.  They also each rely on JPivot, a web interface that lets end users run OLAP queries against Mondrian.  Mondrian actually transforms these queries into standard SQL and thus works well with MySQL, PostgreSQL, Oracle, DB2, and other relational databases.  It uses sophisticated query caching to provide very strong performance.

OLAP schema designer

OLAP schema designer - XML view

OLAP report viewer (JPivot)

OLAP report viewer (JPivot chart)

Conclusion

Jasper has a heavy focus on reporting and analysis.  It has a better web 2.0 user interface than Pentaho and is generally easier to use.  It benefits from better marketing, informational web sites, and documentation.  Overall, it makes researching and finding what users need easier than Pentaho.  The paid version offers ad hoc query support through the web interface (which comes with the open source version of Pentaho) and a dashboard creation web interface.

Pentaho focuses on data integration, ETL, and workflow automation.  As such, it is very powerful.  Today, it has a user interface that is more difficult for business users than is Jasper’s.  The documentation is also less robust than its competition.  On the other hand, it has strong data mining tools and even iPhone integration.  The paid version includes LDAP integration, SSO support, and auditing capabilities (all of which come with the open source version of Jasper).

If your enterprise needs center around report creation, distribution, and business end user self-service, Jasper is probably a good choice.  If your needs are primarily around data integration, reporting automation and workflow, and data mining, then Pentaho is probably a good choice.

Final Thoughts

Open source reporting tools have made great strides in the last two years and they will no doubt continue to get stronger every day.  Competition continues to drive Jasper and Pentaho very hard.  Their capabilities have leapfrogged each other before and it may happen again.  Jasper will grow its abilities in the data integration space and Pentaho will improve its user interface.

Keep watching this space as the players evolve at a pace that only open source can maintain.