here - Overview

Download Report

Transcript here - Overview

ZED Financials
ETL Installation and Technical Detail
Component Structure - Server
From Source
System
To Source
System
Extractor
Transformation
Engine
Including
Translation Engine
And Transformers
Response Translator
Including
Style sheet
Translation
The Response Translator
Is only configured for Source
Systems that require a response
Loader
To Target
System
From Target
System
Configuration - Overview
The ETL configuration is encapsulated in an object referred to as a Server, this includes:

Extractor,

Transformation Engine including Transformer objects,

Loader and

Response Translator
The Server object is multi-threaded to achieve maximum performance but take care when determining the number
of threads to use as too many can result in the JVM spending more time switching between threads rather than
actually running them.
Multiple servers can be run simultaneously so for instance, one server can be running as an on-line service handling
financial trades while another instance operates as a batch process loading CSV files from another source system
etc.
If the source system requires a response from the target system, even if its only pass or fail, a Response Translator
should be configured, obviously if no response is required eg. a batch process, then Response Translator does not
need to be configured.
Configuration – Templates and Maps
The first step is to define the data format provided by the source system and required by the target system, to do
this Templates must be created.

Where either data format is XML, this can be created automatically by using the Tools -> Template Import
feature. To use this feature example files must be placed is specific directories ie. ETL_HOME/template/source
or ETL_HOME/template/target. Currently these are set up for Financial trades sub-divided into Asset Classes
however these can easily be changed as required.

Where source system is a database, the template relates to the columns returned by the SQL Query that
extracts the data, refer to the guidelines for more detail but essentially, aliases should be used for the column
names as this allows for the structure of the query to be altered without affecting the template name.

Where the target is a database, the single table column names should be entered into the template.

Where the source or target is a delimited file, the column position and description should be entered into the
template, the description is only used for info, the column position must start from 0.

Where the source or target if a Fixed Field Width file, there is a screen under the Tools -> Configuration
Assistant tab to help with this.
Once the templates have been defined, the next step is to create a Map which is simply identifying the Source and
Target templates and which fields map to each other.
Configuration – Server
The server object encapsulates the ETL process as outlined below so this section brings it all together.
The important thing to note is that the datatype is the key to how data is processed.

Basic
This section sets the basic environment the server runs in and is fairly self explanatory however the Error logging section needs a bit of
an explanation.
The error logging referred to here is in addition to the normal logging and is used to capture failed XML format messages which can be
modified and resubmitted via the Message mechanic which is available from the Tools menu.

Extractor
This section defines the type and connection properties of the Extractor object ie. the object that is the link between the source system
and the ETL process.
Where appropriate eg. when the source is a database, the operator will be prompted to enter the details of the datatype ie.
 The processing order, applicable when this server processes multiple datatypes from the same source
 The specific datatype
 The format of DATE data, in Java style, examples are given in the Guidelines tab
 The map to be used for this datatype and
 The SQL query filename to be used to extract the data.
Configuration - Server

Loader
This section defines the type and connection properties of the Loader object ie. the object that is the link between the back office (
target ) system and the ETL process. Where the source system requires a response, the Response Translator object is also defined here.
Where appropriate the operator will be prompted to provide details of the datatype ie.
 The specific datatype
 The format of DATE data, in Java style, examples are given in the Guidelines tab
 The name of the file to be generated
When a Response Translator is to be configured, it should be noted that the communication link type can be different from the link
between the Loader object and the Back Office System

Response Translator Stylesheet
When a Response Translator is configured it may or may not require a stylesheet ( Java XSLT ) to be defined, this stylesheet will be used
to translate the response from the back office system format into the format required by the source system, it should be noted that
cpmplex stylesheets can have a performance impact.
Configuration - Server

Datatype Specification
This defines how the datatype of a message is determined. During the configuration the operator will be presented with two different
sets of parameters depending on the type of datasource, these will be covered by the following:
 Source Datatype Field - The field in the Source system data that identifies the datatype of the data
 Datatype - The datatype used throughout the ETL to identify the correct Map etc
 Use Value of Source Datatype Field - If the field exists use the value to identify the identified datatype
 Source Datatype Field has Specific Value - If the field exists, does it have a specific value
 Source Datatype Field has Value Containing Specific Value - If the field exists, does it contain a specific value
 Source Datatype Field does NOT Contain Specific Value - If the field exists, does it NOT contain a specific value
 Source Datatype Field exists a specific number of times - If the field exists specific number of times
 Source Base Elment does NOT Exists - The source does not contain an element whose Base Name is specified
 Source Datatype Field Specific Value or Count or Base Name- If the field exists, the specific value required or the number of times
it exists eg. 1
 Data Map - The map used to transform Source format to Target format data
 Internal Message ID Field - The field in the Mapped Message that contains the unique data identifier supplied by the Source system
 TargetResponse Message ID Field - Where the Source requires a response, this identifies the field in the Recipient system response
containing the unique data identifier
 Target Response Datatype Field - Where the Source requires a response, this identifies the field in the Recipient system response
containing the datatype
NOTE: the Base name of an element is the last sub-element ie. given an element with a name of customer.address, the Base name is
address.
Configuration - Server Transformers

The package ships with a selection of Transformers which are used to massage the message as it proceeds
through the ETL process, the available transformers can be accessed from the ETL Configuration -> Transformers
tab.

Each transformer includes a set of parameters which are applicable.

The transformer list includes two database lookups ie. static table data and SQL generated data, both of these
can be created using the Tools -> Configuration Tools -> Lookup screens.

When creating a server configuration, transformers can be allocated to a particular datatype together with
defining the appropriate values for each parameter from the ETL Configuration -> Server screen.

To add a new Transformer, the operator should create a new Java package to add the Transformer class to and
add the jar file to the ETL Classpath. Source code for a transformer will be made available to be used as an
example.
Configuration – Directory Structure
The ZED Financials – ETL tool uses several pre-defined directories which are sub-directories of the ETL_HOME
directory:

Under the Tools tab there is a facility to import Source and Target system XML format data templates, these
templates should be placed in ETL_HOME/template/source ( target ) directory.

The Response Translator uses XSLT stylesheets, these should be placed in ETL_HOME/mapper/xsl.

Database source extractors use SQL queries, these can be developed anywhere but should then be placed in
ETL_HOME/mapper/data/requests directory.
Limit Manager Installation

As the Limits Manager is provided as a war file there are a couple of edits that need to be made so:
1. Copy the limitManager.war file from the webapp directory to the tomcat/webapps directory
2. If tomcat is running already, it will explode the limitManager.war file into a webapps/limitManager directory, if tomcat
is not running it should be started by running the tomcat/bin/startup script.
3. Edit the webapps/limitManager/WEB-INF/spring/datasource.xml file to set the Oracle database connection details
4. Restart tomcat.
•
To install and configure the Oracle database, from the sql/oracle directory:
1. Start a SqlPlus session connected to the required database schema
2. Create the tables using the command @load_tables
3. Create the stored procedures using the command @load_packages
4. If it is necessary to re-run these commands, the tables can be dropped using the command @drop_tables.
Tomcat Configuration Changes

The default memory sizes available to Java may need to be adjusted, this can be achieved by updating the JAVA_OPTS
property defined in catalina.sh eg.
JAVA_OPTS="-Xmx1024m -Xms512m -XX:PermSize=512m -XX:MaxPermSize=768m -XX:ReservedCodeCacheSize=512m $JAVA_OPTS"

If using Tomcat version 8 there may be cache size warning, this can be rectified by updating the context.xml file by adding:
<Resources
cachingAllowed="true"
cacheMaxSize="100000"
/>
Final Steps

URL for the ETL Configuration Client : http://<hostname>:<port>/lemEtl/index.jsf

URL for the Limits Manager Client : http://<hostname>:<port>/limitManager/login.jsf

The installation will create an empty configuration database, an example configuration is available to pre-load this, to do
this, copy the 2016Jul09_100606 directory from the unzip location to the $ETL_HOME/mapper/backup directory. Once this
is done, the database can be loaded by using the Database/Restore feature of the configuration Client.

If the pre-load from the backup of the configuration database is used, please review all of the setup as things like Extractor
ports, DB connection detail will need to be updated.

During the installation of the Oracle database a selection of users are created, the most important is the admin user, no
password specified so initially, start the Lem client and login as user admin, don’t enter a password and select login. The
user will be prompted to enter and confirm their password. Once logged in as the admin user, review the other users and
create/update as required.

That’s it, ready to rock and roll !!
Pre - Installation Requirements
ZED Fimamcials Limits Manager and ETL is a java based application which requires the following to be installed or
available prior to installation:

Java 6 SDK or later

Apache Tomcat version 6 or later

A directory into which the ETL will be installed, referred to as the ETL_HOME directory

The application uses a Derby database so an options directory must be made available, referred to as the
OPTIONS_DIRECTORY
During the installation OpenJMS and Active MQ can be selectively installed if required, these will be installed into
the OPTIONS_DIRECTORY along with Derby. These can also be downloaded freely from the various sources via the
links provided on the ZED website www.zedfinancials.com
ETL Installation

Confirm the ETL_HOME and OPTIONS_DIRECTORY exist and are writeable by the operator

Download the zip file using the link provided into a temporary folder and unziip.

Run the install script, this will prompt the operator to enter the two required directories and then install the
application into these two specified directories.

As the Configuration Client is provided as a war file there are a couple of edits that need to be made so:

1.
Copy the lemEtl.war file from the webapp directory to the tomcat/webapps directory
2.
If tomcat is running already, it will explode the lemEtl.war file into a webapps/lemEtl directory, if tomcat is not running it should be
started by running the tomcat/bin/startup script.
3.
Edit the webapps/etl/WEB-INF/classes/log4j.xml file to set the limitManagerFileLogger appender to point to the required log
directory.
4.
Edit the webapps/etl/WEB-INF/spring/service.xml file to set the baseDirectory property to be the ETL_HOME and the etlHome
property to be ETL_HOME/mapper.
5.
Restart tomcat.
The Derby configuration database server can be run manually using the scripts provided in the
ETL_HOME/mapper/bin directory but really should be configured by your system administrator to be run
automatically. Before proceeding to the next step, ensure Derby is running.
Configuration

Open a browser window and enter the url http://<hostname>:<port>/lemEtl/index.jsf, this will connect to the
ETL Configuration Client.

Open the Tools/ETL Licence tab and enter the provided licence key into the New ETL Licence Key field and
save.

Open the ETL Configuration/Environment tab, ensure the ETL Home field is correct and select Save, this will
update the directory structure fields.

Before continuing with the configuration the user should be familiar with the Component Structure outlined.