ZNAPZ Blog

Crossover Domains explained

2012-03-14T11:40:00.001+01:00

A crossover domain in Maximo 7 is a value list that copies not only the selected value, but also related data to the record you opened the value list from.

An easy example is the Asset field on a work order. When you select an asset, Maximo also copies the priority from the asset to the Priority field on the work order. This happens because there is a crossover domain for this. Take a look at it: in the application Domains, search for ASSET2WO.

In the Crossover Domain section, you see the object the values will be copied from, also called the Source Object. Actually, this is the same as on a table domain. This source object will be the source of the values for the values list. In this example, the value list will show a list of assets.
If you do not want to show all records from the source object in the value list, you can provide a where clause in the field List Where Clause to filter records.
The field Validation Where Clause is used to specify a where clause that is used against the source object when validating a value that is entered in the originating field. In this example, when you enter a value in the work order asset field, this value will be validated against the Asset object using the validation where clause specified on the domain.

In the Crossover Fields section, you will see the source field, PRIORITY (from the Asset object), and the destination field, WOPRIORITY (from the Workorder object).
You can add additional fields here.

For a crossover domain to work, it has to be linked to the field that has to trigger the crossover action. If you want to copy values from the asset to the work order, you link the crossover domain to the Assetnum attribute on the Workorder object.
You will most likely do this in Database Configuration. However, Maximo out-of-the-box crossover domains are often linked to an attribute through a java class, specified on the attribute. ASSET2WO is an example of that.

There are a number of out-of-the-box crossover domains: ASSET2WO, TICKET2WO, WO2WO, LOC2WO, ...
You can add additional fields to these crossover domains to suite your own configuration.

Modifying size of text boxes

2012-01-02T13:46:00.002+01:00

You probably ran into this problem when using Maximo: Maximo sometimes does not display all the characters in a field, because the text box is too small. Of course, you can solve this. First you need to understand how Maximo determines the display size of the text box. Then you can change these setting and control how large the text box should be.

Understanding text box sizes
Maximo uses the database field size and the settings in fieldsizegroups.jsp to determine the display size of a text box on the screen. In fieldsizegroup.jsp, there is a table containing the different display sizes in relation to the database field size and the data type.
Example:
//For ALN, LOWER, UPPER field types
fieldtypegroup = new Hashtable();
fieldtypegroup.put(new Integer("2"), new Integer("2"));
fieldtypegroup.put(new Integer("30"), new Integer("10"));
fieldtypegroup.put(new Integer("31"), new Integer("40"));

This means that for fields of type ALN, LOWER and UPPER:

the display size is 2 when the database field size is smaller or equal to 2
the display size is 10 when the database field size is smaller or equal to 30
the display size is 40 when the database field size is equal or greater than 31

Modifying settings
To modify the display size of a field, you can do two things: 1) change the database field size so that it falls in a different range, e.g. for a field with an UPPER field type, increasing the database field size from 15 to 35 will increase the display size from 10 to 40. 2) change the ranges and sizes in fieldsizegroup.jsp. Example:

fieldtypegroup.put(new Integer("2"), new Integer("2"));

fieldtypegroup.put(new Integer("20"), new Integer("20"));

fieldtypegroup.put(new Integer("30"), new Integer("30"));

fieldtypegroup.put(new Integer("31"), new Integer("40"));

Changing the fieldsizegroups.jsp ranges will affect the display size of all fields of the specific data type. Changing the database size will affect all the text boxes that are based on that attribute.

Note: Changes to the file fieldsizegroups.jsp will only take effect after you build and deploy the maximo.ear file.

But there is also a trick: if you want to change the display size of only one field on a specific application screen, you can override the settings in fieldsizegroups.jsp. You do this by adding USEFIELDSIZEGROUP="FALSE" to the text box properties in the application xml. Maximo will then use the database field size as the display size.

Example (from asset.xml):

Server names for cluster members

2011-11-10T13:56:00.001+01:00

The name of a server is normally defined by the property mxe.name in the maximo.properties file. In a clustered environment this would lead to all members having the same name since only 1 EAR is used for all these members.

This article will show you how to give each cluster member a unique name.

Maximo will use this approach to define a server name:
1. First Maximo will look for a JVM system property called mxe.name
2. If this is not available Maximo will look for a property called mxe.name in the maximo.properties file

In a clustered environment using the maximo.properties file is not an option. System properties can be set using a JVM command line option:

-Dmxe.name=MxServer003

WebSphere

In WebSphere all the members of a cluster have a unique name and this is stored in a variable called ${WAS_SERVER_NAME}

In WebSphere we only need to do the following:

Servers -> Application Servers -> MXServer003 -> Java and Process Management -> Process Definition -> Java Virtual Machine -> Generic JVM arguments

Add the following to the JVM arguments:

-Dmxe.name=${WAS_SERVER_NAME}

How to validate the server names?

Logfile

We can check the server name in the SystemOut.log. During startup maximo will log a line which looks like this:

[10/25/11 9:59:10:232 CEST] 0000002d SystemOut     O 25 Oct 2011 09:59:10:232 [INFO] BMXAA6450I - Property mxe.name = MXServer003

System Properties Application

In the System Properties application we can check the member names in the section where an instance property can be added. The server name attribute has a lookup list which containes all the different server names.

How labels work in Maximo 7

2011-10-28T14:47:00.009+02:00

How it is stored:

Maximo stores labels in multiple places. When you add a label to a field, tab or section in Application Designer, Maximo stores this in the application XML in the MAXPRESENTATION table and in the MAXLABELS table. For fields, there are also labels stored in Database Configuration, in the title field on the attribute.

How it is loaded:

When an application is opened for the first time after the application server has started, Maximo loads the application XML and gets the labels in this order: first it looks for a label in the MAXLABELS table; if no label is found it looks in the application XML for a label; and finally, if there is also no label there, it looks (only for fields) in the title field in Database Configuration. When this is done, Maximo stores the application definition in the cache memory of the application server.

If your Maximo environment has an application server cluster, every cluster member has its own cache. When the application is changed through one server, only the cache of that specific application server is refreshed. The other servers still use the old version of the application, until their cache is refreshed also by restarting the application server.

How an import is handled:

Importing an application xml in Application Designer causes Maximo to save the application XML in the MAXPRESENTATION table, save the labels in MAXLABELS and clear the cache.

How it works with multiple languages:

You can translate tabs and fields in Application Designer. To do this, select the other language on the login screen and log in. Open the application in Application Designer and edit the labels of the tabs or sections. Maximo stores the translated labels in L_MAXLABELS.

For fields, use Database Configuration to translate the title and description for an attribute while logged in with the other language selected. Maximo stores the translated labels in L_MAXATTRIBUTE.

Oracle escaping special characters

2010-09-07T19:25:00.003+02:00

In a nutshel

Turn of special meaning of chars

-- The ampersand has no special meaning anymore
SET DEFINE OFF;  
SELECT 'V&D' FROM dual;

Escaping characters in LIKE

-- The meaning of the underscore is escaped
SET ESCAPE ON;
SELECT * FROM maxtable WHERE tablename LIKE 'A\\_%' ESCAPE '\\';

Other Examples

Examples - Supressing special chars

-- Special chars are now treated as normal characters
SET DEFINE OFF;  

-- All characters are in the output
SELECT 'V&D' FROM dual;
SELECT 'V\&D' FROM dual;

-- Escape characters works as expected
SELECT * FROM maxmessages WHERE value LIKE '%\%%' ESCAPE '\';

-- No results...there are no tablenames which start with letter A and a backslash
SELECT * FROM maxtable WHERE tablename LIKE 'A\_%';

-- Runs into an error
SELECT * FROM maxmessages WHERE value LIKE '%\\%%' ESCAPE '\\';

Examples - Escape character

-- Set dash (=default) as an escape character
SET ESCAPE ON;
-- or explicitly set backslash as an escape character
SET ESCAPE \;

-- Escape char works as expected
SELECT 'V\&D' FROM dual;

-- Underscore and Percentage are escaped as expected
SELECT * FROM maxtable WHERE tablename LIKE 'A\\_%' ESCAPE '\\';
SELECT * FROM maxmessages WHERE value LIKE '%\\%%' ESCAPE '\\';

-- Both statemenst have the same result (escape char has no influence)
SELECT * FROM maxtable WHERE tablename LIKE 'A_%';
SELECT * FROM maxtable WHERE tablename LIKE 'A\_%';
-- Special chars are now handled as normal characters
SET DEFINE OFF;  

-- All characters are in the output
SELECT 'V&D' FROM dual;
SELECT 'V\&D' FROM dual;

-- Escape characters works as expected
SELECT * FROM maxmessages WHERE value LIKE '%\%%' ESCAPE '\';

-- No results...there are no tablenames which start with letter A and a backslash
SELECT * FROM maxtable WHERE tablename LIKE 'A\_%';

-- Runs into an error
SELECT * FROM maxmessages WHERE value LIKE '%\\%%' ESCAPE '\\';

-- Variable substitution is performed
SELECT 'V&D' FROM dual;

-- Runs into an error
SELECT * FROM maxmessages WHERE value LIKE '%\%%' ESCAPE '\';

Quartz - Run Now

2008-02-22T13:56:00.000+01:00

Quartz is an open source job scheduling system written in Java. Executing a job in quartz is based on the definition of triggers. But what if you are in a development environment where you don't want to schedule jobs and want to be able to run them on demand? The API of version 1.6.0 tells us that the triggerJob method in the Scheduler class can be used for that. Great! But the triggerJob method also seems to throw away your job from the scheduler. So the second time you run the job it's bingo!

A possible workaround we figured out for this is to reschedule the job with a simple trigger which will run immediatly:

scheduler.deleteJob(jobName, groupName);
scheduler.scheduleJob(
    jobDetail,
    new SimpleTrigger(jobName, groupName)
);

You could of course also use the rescheduleJob method but then you have to be sure the job is already scheduled. In our situation this was not the case.

Sparklines with the Google Chart API

2008-02-14T16:53:00.003+01:00

Sparklines are described in Edward Tufte's book Beautiful Evidence (2006) as “small, high resolution graphics embedded in a context of words, numbers, images”. In one of our projects we needed sparklines and one option was to use this open source package by Larry Ogrodnek. Just out of curiousity I was wondering how far we would come if we would use the Google Chart API.

That was until I ran into this article where I saw sparklines created with the Google Charts API and without the x- and y-axis. How the hell did he do it?

Well ... it seems to be an undocumented feature also used by Google on their financial site. Instead of using cht=lc (chartype is line chart) you can use cht=lfi (charttype is financial line chart).

Grab the code:

http://chart.apis.google.com/chart?chs=80x20&cht=lfi&chco=0077CC&&chm=B,E6F2FA,0,0,0&chls=1,0,0&chd=t:27,25,25,25,25,27,100,31,25,36,25,25,39,25,31,25,25,25,26,26,25,25,28,25,25,100,28,27,31,25,27,27,29,25,27,26,26,25,26,26,35,33,34,25,26,25,36,25,26,37,33,33,37,37,39,25,25,25,25

Lucene uses Google

2008-01-29T17:27:00.001+01:00

Today I was figuring out how to integrate the Hibernate search API in one of my projects. At a certain point the docs referred to the Lucene documentation. My old pal Google provided me the link to the Lucene homepage and when I opened it up I saw something very weird in the upper right corner: "Search the site with Google"

Yep, Lucene seems to be using Google as a search engine! At first I was a bit astonished about it. But why bother about it, the Google service is out there and it just works. After all Lucene is a great tool and I hope to tell you more about it in the near future.

PDF documents with iText

2007-05-16T07:32:00.000+02:00

Just ran into a tool called iText which is focused on creating PDF documents. Read more about ithere. A step-by-step guide to iText is available here at IBM or at O'Reilly

iText is used by BIRT an eclipse based reporting tool

Unix: finding big folders

2007-02-27T12:26:00.000+01:00

How to find the folders which are consuming much diskspace. It's easy using du:

du -k | sort -nr | pg

And the bigist folders will be on top of the list.

Unix Performance Management

2007-02-23T06:56:00.000+01:00

Today I was with one of our customers and we ran into a performance issue. The first request to a remote procedure call took 15 seconds extra time. Uptill now we could't figure out wat the cause was of this performance problem but these two tools could be very usefull for measuring things on OS level.

Viewing the memory statistics every 5 sec:

vmstat 5

Disk I/O statistics every 5 sec:

iostat -Ds 5

Sun Solaris

2007-02-21T10:36:00.000+01:00

Which version of MySql do I need to download for this sun solaris machine? To get an answer to this question the following commands can be very usefull:

Revision info for hard- and software:

showrev

Detailed processor info:

prsinfo -v

Instruction set supported by the machine:

isainfo -v

Spaces in filenames

2006-11-21T19:10:00.000+01:00

Today's desktop users are familiar with filenames including spaces. The hell begins when those users can store files on a unix server using a windows desktop.This morning I had to change the permissions of a bunch of files on solaris. Normally i would just type in the command and run it .... but NOT this morning. A lot of files which i had to access contained spaces pffff. The unix shell is very powerfull ... so I knew there was a simple solution. I am just writing it down here so you can use it in the future:

find . -type d | sed -e 's/ /\\ /g' | xargs sudo chmod 777

Google maps

2006-10-13T13:40:00.000+02:00

My first steps in the world of Google API's and services.A few weeks ago google anounced the complete coverage of the Netherlands. Now I would be able to translate a street address into a longitude/lattitude pair. The last barier for not using the maps api in my application was gone.It was amazing how quickly I had my first integrated map. Most of the effort was not in learning the api and using it but finding the right hooks to integrate it in our cms viewbox. A first draft of this integration is now being used as a demo in our office. Curiously I am waiting for all the brilliant ideas that will pop up the next week.