Tuesday, 14 January 2014

Performance testing oracle JD Edwards EnterpriseOne JDE

I’ve done a lot of performance testing of EnterpriseOne.  Back in time I was involved in testing over 10000 users with load runner on a mix of web and citrix.  I’ve now started doing a lot of specific JD Edwards load testing using Oracle Application Testing Suite (OATS). 

Of course OATS can test all of the JD Edwards suite natively, because it’s oracle.  But it can also natively test all of the oracle product suite natively (in fact – ANY web based application).  OATs has accelerators for many of the oracle products, which make recording and reading the scripts a breeze.

Load testing itself is not overly complicated and you can execute this as a “one off” without any problems.  But how do you know what is normal? What is fast? Where improvements can be made? It’s difficult when you only have one set of performance data to use for your comparisons.  What I mean in this situation is, if you are getting average response times of .75 sec, is this good for a 400 user concurrent test?  Can you improve that?  Is 50% CPU utilisation on the database server normal with 300 users connected?  All of these questions are difficult to answer without a frame of reference.

At Myriad, we do load testing for lots of clients with lots of different architectures and we can tell what is fast and what is slow at the various application tiers.  We know where you should be putting your horsepower and we know what is going to make the best difference for your ERP performance.

So what I guess I’m saying is that you can have all of the findings and metrics in the world, but the interpretation of these results is where the value is.  I guess there is a similar analogy with big data, you can have all of the data in the world, but it means nothing until you’ve interpreted it.  It seems to me that the real value of metrics is found with interpretation & comparison rather than merely representation.  In the simplest form, your information is valuable as soon as you make a change and then compare the results.  The metrics start to have value with comparison.  This is true with your performance tuning data, you need to compare your previous results to know if you are making improvements or not, but unless you compare with industry standard benchmarks, you won’t know if there are other gains that you can easily make.

The best thing to do is create a benchmark. Then any changes you make in the system can be compared against this benchmark.  You’ll be able to quantify any performance changes with the user community by comparing the results with your known benchmark.  This could be done for a new tools release, a new piece of hardware, network architecture…  Anything.

Our LTAAS (load testing as a service) offering can enable you to do this remotely, any time you like.  We are able to engineer specific tests for your site and execute them when you want in the concurrency that you want.  We are also able to configure all monitoring to ensure that not only the performance data (in terms of response times) is stored, but also the performance metrics from the hardware.  We can save off your results and then compare them if you execute the tests again.  We are also able to compare these results with know industry benchmarks to ensure that your performance profile is within industry standards.

Remember that you can do all of the tuning in the world, but 1 bad query from a user can bring the system to it’s knees!

This is a great toolset that you are able to augment with your own bag of tricks.  Send me an email shannon.moir@myriad-it.com if you’d like to know more.

windoze search is a COMPLETE DONKEY

Tell me this, please…

I do the following in windows explorer search:

image

contents:ORA and I get NO results

I search for ORA and then click file contents and I get no results

image

I then go to my trusty command window and enter:

D:\JDEdwards\E910\log>findstr ORA *.log

and I get a TON of results.

I honestly don’t know how we are supposed to trust an OS (a commercially focused OS like 2008R2) if it cannot search the contents of TXT files / log files “out of the box”.

Do I really need to change the role of the server to enable windows search service to search the contents of log files?

Wednesday, 4 December 2013

JD Edwards performance testing using Oracle Application Testing Suite

I thought that I’d provide some very generic guidelines for performance testing and JD Edwards 9.1.  These figures can be used as a guide for your systems performance.  If course, these tests and scenarios are very limited.

I have a scenario, 1 web server, 1 DB server, 1 app server.  I log into E1, I launch AB, index search 1 record, close AB, open BV, run R0010P, close BV logout close browser.

I create a load test using Oracle Application Test Suite.

I configure all of my counters

I configure 20 concurrent users (to being with)

I choose to have 0 wait time, so I’m perf testing.

I run this scenario for 10 minutes.

What do you think the results might be? 

Name Value Min Max Avg
Transaction/s .827 .082 2.579 2.272
Pages/s 8.8 .667 26.2 22.68

image

 

image

 

So these are very interesting stats.  Each user is getting 1 page per second, Each user is getting 10 hits per second.  Each user is generating 58KB/s.

I find the final item interesting, as this allows you to know exactly how much a very fast moving user is using on average for E1 – this is 0 wait time, so a user could not actually achieve this (none that I know).

Like I said, this is a 1 off very simple sample.

Oracle Application Testing Suite is fantastic and is exceptionally easy to use.  Although, what you do not have (and I do) is the ability to compare values between sites and know where you can make considerable performance improvements!

Wednesday, 27 November 2013

Oracle procedures and calling them from a BSFN

We are doing some pretty edgy stuff at the moment.  We are getting JD Edwards to call oracle procedures via OCI calls in kernels.  This is some nice stuff to have complete commitment control over function and procedure calls from within E1.

We are working on all of the syntax and semantic issues at the moment.

At the end of the day we are going to have a generic BSFN that is going to take 3 parameters:

procedurename

parameters in (pipe delimited)

parameters out (pipe delimited)

Therefore, the BSFN can be nice and generic and send a bunch of parameters to the stored proc, it’ll extract them all with it’s beautiful PL/SQL logic and hand any returns back when it is done – nice!

Note that this will be timed out if there is no response to the BSFN in 60 seconds, which will be no good for interactive apps

create or replace Procedure TESTDTA.ProcTest
(procName IN varchar2, environ IN varchar2, paramList IN varchar2, resultsList OUT varchar2)
AS
  nextID number;
BEGIN
DBMS_OUTPUT.PUT_LINE('In Procedure');
SELECT testdta.ProcedureSeq.nextval INTO NextID FROM DUAL ;
insert into TESTDTA.ProcedureTest values (ProcTest.procName, ProcTest.environ, ProcTest.paramList, CAST(nextID as varchar(20) ) );
END;

Run it with the following:

declare
  ReturnVal  varchar(20);
begin
  TESTDTA.ProcTest('This is procedure name', 'DV910', 'This|Is|999|params', ReturnVal);
  DBMS_OUTPUT.ENABLE(1000000);
  dbms_output.put_line(ReturnVal) ;
END ;

Check that things get inserted. 

select * from testdta.ProcedureTEst;

image

Nice, all is working. Note that the procedure needed to be created in the same schema as the sequence and the table.  If I did not create the procedure in the same owner / schema – it was unable to find the sequence or the table even when I explicitly referenced them.

So, this is perfect.  I have a working procedure in my TESTDTA owner – now I have to code the OCI routines to call the procedure.  There are two ways of doing this:  OCI (nice, neat secure) or system out a call to a vbscript (not nice, neat or secure – but quick).  I’m going with OCI.

OCI is going to be slightly painful, for a number of reasons.  I need to use 32bit OCI libraries and I also need to do “loadlib” calls and function pointers – as the oci libs are not going to be in my lib path when compiling BSFN’s on the server (or the client).  This is somewhat painful.  I’ll get it all working standalone and then build my function pointers into a BSFN.

This might be a longer post than usual, so I’ll do it in parts.

Thursday, 21 November 2013

Promoting E1Pages not working–Work with LCM activity Rules

This is a really quick post, because I know that I’m going to forget it again.  If you are having issues promoting e1pages using P982400, you need to ensure that you have all of the security set up correctly in P982405.  This application controls who can promote e1pages (or user defined content) from one status to another.

  • Pending Review
  • Rework
  • Pending Approval
  • Approved

The above are sample from and to status that you control with this app.

Remember also that these tables are stored in the central objects data source, so any environment associated with the pathcode will see the e1page

clip_image002

Desktop BIP Installation

Getting issues with 64 bit / 32 bit combinations of office and windows?  This is based on 64 bit windows 7 and 32 bit office 2010, yes it can work!  The tricky bit is getting the version of java right for the installer.

Preparation

Close ALL office products (word, excel, powerpoint and email)

Assumption you are running 32 bit office and 64 bit windows

Download 64 bit java and install

http://java.com/en/download/manual.jsp This page, has the 64 bit download. Download and install this.

http://javadl.sun.com/webapps/download/AutoDL?BundleId=81821

clip_image002

Got control panel and type java, select it

clip_image004

Goto java tab, then view button

Ensure that a version in “Program files” is selected (this is the 64 bit location), not the version in “Program files (x86)

clip_image005

The above is 64 bit!

Install BIP desktop

Now run the bin installer \\aupdc00-jds01d\E910\BIPDesktop

Right click the exec and run as admin

clip_image006

clip_image007

Choose run

clip_image008

Ok

clip_image009

clip_image010

It’ll be done

clip_image012

Customising the login for JD Edwards 9.1 and above.

 

Change the page title

Edit this file:

D:\Oracle\Middleware\user_projects\domains\E1_Apps\servers\JWB01D_DVCMSAU_81\stage\JWB01D_DVCMSAU_81\app\webclient.war\share\Html4login.jsp

Change this

<title>"JDE CMS DV AU"</title>

<webgui:pagehead addtitle="false"/>

Note that this is really handy for favourites too!

clip_image002

Then change the icon’s and text – Just for the login page. Replace the oracle logo change the text from “JD Edwards EnterpriseOne”

Note that you need to create the oracle_logo2.png

File to edit:

D:\Oracle\Middleware\user_projects\domains\E1_Apps\servers\JWB01D_DVCMSAU_81\stage\JWB01D_DVCMSAU_81\app\webclient.war\share\loginDecoration.jspf

Changes:

if ( (companyLogoImageURL != null) && (companyLogoImageURL.length() > 0) || smallScreen )

{

if (companyLogoImageURL.length() != 0)

{

sb.append("<img src=\"");

sb.append(resourceUrlBuilder.createUrl("/share/images/rcux/oracle_logo2.png"));

sb.append("\" ALT=\"ORACLE &reg;\"");

sb.append(" width=\"219\" border=\"0\" height=\"75\">");

}

}

Then also change

sb.append("</td>");

sb.append("<td width=\"100%\" class=\"loginValignTop topbody bottomborder\">");

sb.append("<span id=\"appName\" style=\"position: absolute; top: 25px; left: 220px;\" class=\"appname\"\" > xxxx CMS Australia Development </span>");

sb.append("</td>");

See I changed the height and the title for the login only. Note that the FLAG and the workd “CMS DEV” are part of the image oracle_logo2.png. The words “XXX CMS Australia Development” are the second change that is referenced above.

clip_image006

Job is done. You now have a modified login page and no others.

This can be seen immediately with a refresh. You need to make the changes on the server (of course)

Change colours of top navigation bar:

clip_image009

\\aupdc00-jwb01d\d$\Oracle\Middleware\user_projects\domains\E1_Apps\servers\JWB01D_DVCMSAU_81\stage\JWB01D_DVCMSAU_81\app\webclient.war\share\webguistylesheet.jsp

.topnavtable {

background-color: #006600;

padding-top: 0.2em;

height: 2.5em;

white-space: nowrap;

}