原文:https://github.com/OSBI/mondrian/blob/4.0/RELEASE.txt
Mondrian 4.0.0 beta release notes
=================================
Introduction
------------
Mondrian 4.0.0 is a beta release of a new major version of the
leading open-source OLAP engine. To find out more about mondrian,
go to http://mondrian.pentaho.com.
Contents
--------
The distribution is contained in the file mondrian-<version>.zip.
Each distribution contains the full Mondrian source code, as
mondrian-<version>-src.zip, documentation including generated API
documentation.
The main interface to Mondrian is its olap4j driver,
mondrian.olap4j.MondrianOlap4jDriver.
This release runs on Java version 1.5 and higher. For
backwards compatibility on Java 1.4, substitute
mondrian-jdk14.jar for mondrian.jar, and add
retroweaver-rt-1.2.4.jar to your classpath.
For further installation instructions, see
http://mondrian.pentaho.com/documentation/installation.php
or the included file doc/installation.html.
Main functionality in this release
----------------------------------
- Measure groups
- Attributes
- Time dimension generator
- Composite keys
- Physical schema
- API to define aggregate tables (not yet complete)
- Schema format does not require XML elements to be in a
particular order.
- Schema validation gives multiple errors and warnings, and
errors and warnings have a precise XML location.
- If connection property 'Ignore' is specified, warnings do not
prevent a connection from being established.
Limitations
-----------
Some functionality is not complete in this beta version.
- Ragged hierarchies are not supported.
- No all native SQL optimizations supported in Mondrian-3.x
still work.
- There are approximately 250 test failures out of 2700 tests.
- Pentaho Analyzer does not yet work against Mondrian-4.
- It's not as easy as it should be to download and load sample
data sets such as FoodMart.
- Default value of Attribute.hasHierarchy is currently false.
Before production, this will change it to true, if attribute
is not included in any hierarchies.
API and MDX language changes from 3.x to 4.0
--------------------------------------------
- Mondrian-4 is compatible with olap4j version 1.0.1.
- In MDX, hierarchy elements must be specified using
[dimension].[hierarchy], and level elements must be specified
using [dimension].[hierarchy].[level]. (In Mondrian-3.x, this
behavior was enabled only if the property
mondrian.olap.SsasCompatibleNaming was set to true.)
- The following deprecated APIs have been removed:
-- mondrian.olap.DimensionType
-- mondrian.olap.LevelType
- createDimension has been removed from the API mondrian.olap.Schema
- TODO
Other changes from 3.x to 4.0
-----------------------------
The XML schema format is not compatible between
versions. Mondrian-3.x software would not be able to read a
Mondrian-4 schema, and Mondrian-4 software can read a Mondrian-3
schema only because it recognizes that the schema is a legacy
format and internally converts to Mondrian-4 format.
We recommend using the metamodelVersion attribute of the Schema
element, to make it clear which version the schema is intended
for. This attribute is mandatory from 4.0 onwards.
The schema converter can handle most schemas, but has the
following limitations. It cannot recognize uses of columns in
expressions and convert them to Column elements.
The VirtualCube element is obsolete. The equivalent in Mondrian-4
is a Cube that contains multiple MeasureGroups.
The AggName element is obsolete. In Mondrian-4 you define
aggregate tables using the MeasureGroup element (which is also
used to define fact tables).
The AggPattern element is obsolete. There is no direct way in
Mondrian-4 to define a set of aggregate tables by
pattern-matching the names of tables and columns in the
schema. You can use the new aggregate table API to scan the
schema and define aggregate tables.
The standard demonstration schema is now called
FoodMart.mondrian.xml. (The longer extension helps tools such as
Pentaho more easily identify the purpose of the file.) It is in
version 4 format, of course. The previous demonstration schema,
which was called FoodMart.xml in version 3, is now called
FoodMart3.mondrian.xml.
In 3.x releases, the mondrian-<version>.zip file contained the
FoodMart data set in MySQL and Microsoft Access formats, and
there was an additional mondrian-<version>-derby.zip containing
the FoodMart data set in Apache Derby format. The binary
distribution of this release does not contain the data set. It
can be downloaded from
http://repo.pentaho.org/artifactory/repo/pentaho/mondrian-data-foodmart.
JPivot is no longer included with the release. Neither is there a
mondrian.war file in the distribution.
Schema workbench is obsolete and has been removed. Pentaho hopes
to build a replacement, but no firm plans exist at this time.
XMLA server has been removed from the Mondrian code base. It is
now a separate project. See
https://github.com/julianhyde/olap4j-xmlaserver.
The "high-cardinality dimensions" feature has been removed. There
are better ways to support dimensions that have large numbers of
members.
See CHANGES.txt for a list of all source code changes since the
previous release.
Upgrading from 3.x to 4.x
---------------------------------------
Mondrian 3 schemas will be automatically upgraded to Mondrian 4
schemas internally. Set the log4j log level for
mondrian.rolap.RolapSchemaUpgrader to DEBUG to log the XML for the
upgraded schema.
Bugs and feature requests fixed for 4.0
---------------------------------------
TODO
Removed
-------
To be removed
-------------
Property.formatter (XML attribute)
SchemaGrant.access value "all_dimensions"
# End
相关推荐
mondrian4 schema设计说明文档,压缩包解压后打开index.html。4版本,官网链接 https://mondrian.pentaho.com/head/documentation/schema.php,无法访问,直接用该资源即可。附3版本地址: ...
mondrian和olap4j是两个在大数据分析和在线分析处理(OLAP)领域中重要的开源工具。Mondrian是一个关系型多维数据集引擎,它允许用户通过SQL查询复杂的数据集,而olap4j则是一个Java API,用于与各种OLAP服务器交互...
mondrian支持使用Log4j等日志框架,可以通过修改log4j.properties文件来设置日志级别。 7. **使用mondrian说明文档**:提供的说明文档会详细介绍mondrian的功能、API使用以及配置细节。仔细阅读并理解这些文档,将...
4. **功能介绍** - **多维数据模型**:Mondrian允许用户定义多维数据模型,包括事实表、维度表、层次结构和成员。 - **MDX查询**:Mondrian支持MDX(多维表达式)语言,这是一种强大的查询语言,用于从多维数据...
4. **Hierarchies and Levels**: 层次是维度内的组织结构,而级别是层次中的具体节点。例如,时间维度可能有年、季度、月的层次,每个层次又有相应的级别。开发者可以通过`Hierarchy`和`Level`类来构建这些结构。 5...
[Manning Publications] Mondrian 实战 英文版 [Manning Publications] Mondrian in Action E Book ☆ 出版信息:☆ [作者信息] William Back D Nicholas Goodman Julian Hyde [出版机构] Manning ...
《联机分析OLAP之Mondrian:深入解析与应用》 OLAP(Online Analytical Processing),即在线分析处理,是数据库技术中的一个重要概念,它主要用于支持决策者的数据分析需求。在OLAP系统中,Mondrian是一个开源的...
4. **数据源和连接**:mondrian支持多种数据源,包括JDBC连接。配置数据源涉及到定义数据仓库的物理结构和逻辑视图。 5. **XMLA连接**:XMLA是一种标准协议,允许客户端应用程序与OLAP服务器进行通信。mondrian源码...
4. **wsdl4j-1.5.1.jar**:WSDL4J是Web服务描述语言(WSDL)的Java实现,它允许处理和操作WSDL文档。在mondrian XMLA环境中,这个库可能用于生成或解析描述mondrian服务的WSDL文件,使得客户端能够了解如何正确地...
4. **处理结果**:解析返回的结果集,将其转化为易于理解的格式,如表格或图表。 **四、使用Java与Mondrian交互** 在Java环境中,你可以使用JDBC驱动来连接到Mondrian服务器。首先,需要在项目中添加Mondrian和...
mondrian是开源的在线分析处理(OLAP)引擎,它主要被用来构建数据仓库和商业智能系统的报表功能。在Web项目中,Mondrian通常与Pentaho Data Integration(Kettle)、Pentaho BI Server等工具结合使用,提供强大的...
自已刚开始学mondrian,里面有两个文件,一个是mondrian自带的一个演示示例,是基于oracle实现的过程。还有一个是eclipse中配置mondrian源代码的过程。都是我自己总结的,原创。QQ:6855957
Mondrian 源码分析和改造设计 Mondrian 源码分析和改造设计是 OLAP(Online Analytical Processing,联机分析处理)服务器的源码分析和改造设计。Mondrian 是一个基于 Java 的开源 OLAP 服务器,提供了多维数据分析...
- **推荐版本**:`mondrian-3.0.4.11371.zip` **2. 部署Mondrian** - **服务器环境**: - 操作系统:Windows 2000 - Web 服务器:Tomcat 5.5.20 - 关系数据库:MySQL - **新建foodmart数据库**: - 下载 `...
【Mondrian是什么】 Mondrian是一个开源的Java实现的在线分析处理(OLAP)引擎。它使用多维表达式(MDX)语言来查询数据,这些数据来源于关系数据库管理系统(RDBMS)。Mondrian的设计允许开发者将其轻松地集成到...
这个包可以帮助开发者确保Mondrian的olap4j实现与其他符合标准的工具和应用兼容。 5. **eigenbase-xom-1.3.0.11999.jar**:Eigenbase XOM(eXtensible Object Model)是用于解析和生成XML文档的库。在Mondrian中,...
4. **前端展示**: mondrian提供了RESTful API,可以与各种前端报表工具(如Tableau、JasperReports等)集成,或者你可以自定义JavaScript客户端来交互式地展示分析结果。 **Maven插件** 在mondrian-war-maven-...
mondrian是一款开源的数据分析工具,常用于构建在线分析处理(OLAP)系统,尤其是在Web环境中。本教程将深入探讨mondrian的使用,包括如何创建Cube、维度、度量以及相关属性,同时也涵盖了MDX(多维表达式)的语法...
4. **示例分析**: - 在提供的 XML 示例中,"Sales" 数据立方包括了 "Time" 和 "Gender" 两个维度,以及 "Unit Sales" 和 "Store Sales" 两个度量。"Time" 维度由 "Year"、"Quarter" 和 "Month" 三个层次组成,而 ...