Best practice to initialize my instruments

I wonder what is the best practice to initilize instruments on test stand.
i have a vi that will ready a INI configuration file and map my instruments and will output a cluster(DMM, Scope, etc..)
so when starting the sequence on pre UUT i will load that ini file and pass it to some local variables.(dmm, scope, function generator etc..).
the question is here. is it better to initialize every instrument on the sequence one by one, or is it better to make a vi and initialize everything on one vi.
or do a call to a sequence where i initialize all step by step?
just some random thoughts but i wonder what is the best practice.

I initialize each instrument in a special sequence in TestStand. Seems a little easier to maintain my drivers that way. It is also a little cleaner than doing a state machine or sequence everything with an error cluster in LabVIEW.
Of course, I also have each instrument with its own routine to read the configuration file and initialize the driver/instrument. So I just have to pass the configuration file around inside of TestStand.
There are only two ways to tell somebody thanks: Kudos and Marked Solutions
Unofficial Forum Rules and Guidelines

Similar Messages

Best practice for initialization

Hi,
this must have been covered hundred of times before - nevertheless I didn't find a good answer.
My webapp has to initialize some objects, e.g. load properties from a file, init log4j etc. What's the best place to put these initializations into (static fields and static initialization in one (but which ?) of the servlet classes - or use a singleton, but how to make sure that it's loaded first).
How can I control the sequence in which the classes will be loaded? (Please don't refer to newly introduced features in servlet 2.3 spec - our app server doesn't support it :-(
Thanks,
Stefan

A singleton class would be a good solution. If you do that, it does not matter which servlet is loaded first; the first one that accesses the class will create the object that does the initialization, and subsequent ones will just use that same object.

Which wayis the best practice ?

Hi ,
I am doubt of which way is the best practice to initialize instance component in a Panel.
ie .is there any diference, in terms of performance , between initializing component inside the constructor or just outside of constructor?
1)
public class MyPanel extends JPanel
private static JScrollPane scrollPane = new JScrollPane();
public CustomerTaskPane()
2)
public class MyPanel extends JPanel
private static JScrollPane scrollPane = null;
public CustomerTaskPane()
scrollPane = new JScrollPane()
}

Correction for the above reply.
*btw can i avoid from declaring the static field of a class itself when i need to implement the class which follows Singleton pattern like.
public class CustomerMainPanel extends JPanel
public static CustomerMainPanel customerMainPanel = null;
public static synchronized CustomerMainPanel getInstance()
if( customerMainPanel == null )
customerMainPanel = new CustomerMainPanel();
return customerMainPanel;
}

Teststand best practices architecture for communicating with TCP instruments

Hi,
I am using Teststand 2014 and LabVIEW 2013 SP1 as the module adapter.
My Teststand sequence file includes communication with 3-5 different instruments via TCP/IP.
Such as Spectrum analyzer, generator and other RF measurement devices.
There are many steps located in subsequences that implement code to communicated with these devices whether to fetch , query or write.
What is the best architecture to implement this program ?
As I can see it there are many possibilities, such as :
1. Launching a subsequence as a new thread in the Sequence Setup. This subsequence calls a VI that dynamically register for events/queue that can be launched from any step. (in this case there is a VI running in the background that can perform Write/Read/Qurey to the instrument and he is triggered from any step using an event or dequeue element)
This option can be duplicated for any instrument or only once for all of the instrument with prioritizing the event queue.
2. Creating a communication reference in the Setup of the Main Sequence. And passing it to any step that need it. (there is also the question how to pass this reference… in Teststand globals or in LabVIEW queue).
another 2 small questions that I encountered -
Is using LabVIEW queues and notifiers in TS steps and obtaining their reference (in the LabVIEW code) by name is considered best practices ?
If I would like to run only one VI in a new thread, should I use a new subsequnce and set it as a new thread or use the Run VI Asynchronously ? what are the differences ?

I use Action Engines to hold my VISA Sessions and just write my LabVIEW VIs to use those to send commands and read data. Any returned data can be analyzed in the VI or passed on to TestStand as a step result. You just need a sequence to run at start up to initialize your Action Engines and another one to close them.
There are only two ways to tell somebody thanks: Kudos and Marked Solutions
Unofficial Forum Rules and Guidelines

Best Practice for Designing Database Tables?

Hi,
I work at a company for tracking devices (GPS Devices). Our SQL Server database is designed to have a table for each device we sell, currently there is 2500 tables in our database and they all have the same columns they only differ in table name. Each device
sends about 4K records per day.
currently each table hold from 10K records to 300K records
What is the best practice to design a database in this situation?
When accessing database from a C# application, which is better to use, direct SQL commands or views?
a detailed description about what is best to do in such scenario would be great.
Thanks in advance.
Edit:
Tables columns are:
[MessageID]
,[MessageUnit]
,[MessageLong]
,[MessageLat]
,[MessageSpeed]
,[MessageTime]
,[MessageDate]
,[MessageHeading]
,[MessageSatNumber]
,[MessageInput]
,[MessageCreationDate]
,[MessageInput2]
,[MessageInput3]
,[MessageIO]

Hello Louis, thank you so much for your informative post. I'll describe in detail what situations I came through my 9 months of work in the company (working as a software engineer, but I am planning to take over database maintenance since no one is maintaining
it right now and I cannot do anything else in the code to make it faster)
At every end of the month our clients generate report for the previous month for all their cars, some clients have 100+ cars, and some have few. This is when real issue start, they are calling their data from our server through internet while having 2000
unit sending data to our server, they keep on getting read time out since SQL Server gives priority to insert and hold all select commands. I solved it temporary in the code using "Read Uncommitted" once I initialize a connection through C#.
The other issue is generating reports for a month or two takes lots of time when selecting 100+ units. Thats what I want to solve, the problem is the one who wrote the C# app used hard coded SQL Statements
AND
the company is refusing to upgrade from SQL Server 2003 and Windows Server 2003.
Now talking about reports, there are summary reports, stops reports, zone reports ..etc most of them depend usually on at least MessageTime, MessageDate, MessageSpeed, MessageIO and MessageSatNumber.
So from your post I conclude that for now I need to set snapshots so that select statements don't get kicked out in favor for insert commands, but does SQL Server automatically select from the snapshots or do I have to tell it to do so?
Other than proper indexing what else I need? Tom
Phillips suggested Table partitioning but I don't think it is needed in my case since our database size is 78GB
When I run code analysis on the app, Visual Studio tells me I better use stored procedures, views than using hard coded Select Statements, what difference will this bring me when talking about performance?
Thanks in advance.

Best practice for responding to keyDown event

I am writing my first Cocoa application. It's a simple utility app -- a calculator. I've got everything working, the last thing I want to add is I want the app to respond to key press events. So, you could click the "4" button with the mouse, or you could just press the "4" key on the keyboard. I'm reading all the event handling documentation, but I'm not picking up on the recommended approach for getting an NSResponder (or NSView) that handles the event.
I can make a subclass of NSWindow and tell the interface builder to make the main window of that type. Then when I override the keyDown message in my new NSWindow subclass, it gets the events, but I'm not sure how to connect it to my application delegate class (since I basically just want to switch on the key pressed and call the corresponding message in the delegate that the button "select" action targets.
Or, I think I heard something about doing this by changing the super class of the application delegate class from NSObject to NSView, or NSResponder or something. I haven't poked at that idea much, but I'm not sure how to actually get the application delegate object then actually set up as the first responder (I think that's what I'd want to do).
So, I'm hoping someone can give me specific directions on how/where I should implement my keyDown method, but I'd also like to understand what the recommended "best practice" is for doing this. What would Apple say is the right architecture for doing this? I'm as much interested in learning how to make this work as learning the Apple design pattern behind it.
The structure of my program is as follows:
I have one nib file (actualy a .xib) where I've designed my window with a textbox for display and a hand-ful of buttons. I've registered outlets and actions from there to the AppDelegate class (the one that the standard project template sets up for you). Following the MVC pattern, I then have a Calculator class that the AppDelegate initializes and all the actual logic of the calculator is in that class. The AppDelegate basically just responds to messages from the buttons, passes them along to methods in the Calculator class, and then asks the Calculator for the current display value and updates that back into the textbox. So, the graphical elements in my nib file are the "view", the AppDelegate is the "controler" and the Calculator class is the "model". Right?
So, who is supposed to be handling keyDown events? Seems like the view should capture them and send them to the controller, but how do I set that up?

Dear Noppong,
You can do it with multiple ways like :-
1. Get the current host name and make complete URL with using host name for the webdynpro iview.
request = (IPortalComponentRequest) this.getRequest();
HttpServletRequest req = request.getServletRequest();
StringBuffer strURL = req.getRequestURL();
2. Create the KM Document or Link for webdynpro Iview OR Create the WPC Web Page for the webdynpro ivew
Refer to [http://help.sap.com/saphelp_nw70/helpdata/en/06/4776399abf4b73945acb8fb4f41473/frameset.htm|http://help.sap.com/saphelp_nw70/helpdata/en/06/4776399abf4b73945acb8fb4f41473/frameset.htm]
[http://help.sap.com/saphelp_nw70/helpdata/en/06/4776399abf4b73945acb8fb4f41473/frameset.htm|http://help.sap.com/saphelp_nw70/helpdata/en/06/4776399abf4b73945acb8fb4f41473/frameset.htm]
[http://help.sap.com/saphelp_nw70ehp1/helpdata/en/ff/681a4138a147cbabc3c76bde4dcdbd/content.htm|http://help.sap.com/saphelp_nw70ehp1/helpdata/en/ff/681a4138a147cbabc3c76bde4dcdbd/content.htm]
Hope it will helps
Best Regards
Arun Jaiswal

Best practice for declaring and initializing String?

What is the best practice for the way Strings are declared in a class?
Should it be
private String strHello = "";
or should I have the initialization in the constructors?

The servlet constructor is usually called once, when the servlet is first accessed. But then again maybe something else happens, google servlet life cycle if you must know.
But let's take a step backwards here. It seems like you are trying to put fields into servlets. Don't do that. When two users fetch the servlet's URL at the same time, the fields are shared between the two hits. If you store something like HTTP parameters in the fields, the two hits' parameters will get mangled. The hits can end up seeing each other's parameter values.
The best way is not to have fields in servlets. (Except maybe "static final" constants, sometimes rarely something else.) Many concurrency worries go away, servlet life cycle worries go away, servlet constructors go away, init() usually goes away.

JEE5 Application Architecture Best Practice.

Hi Everybody
I am going to redesign a moderate size application (not v big but larger then normal).
Now I have few Question in my mind.
I am using JSF as front-end, EJB3 Session Bean for Business Logic and last but not the least JPA as domain model.
1 - With JPA we have a domain classes. Now its better to use entity as manage-bean for JSF or manage bean should be saperate.
2 - Using DTO (Data Transfer Object) is good practice or not in JEE5.
3 - Simplicity or Complexity but with EntityManager I feel no need of DAO but I am used to with DAO pattern. So again as best practice I have to make 1 session bean as DAO and call it from all the session bean where I write business logic or forget about DAO session bean and call EntityManager from all session bean everywhere.
4 - For initializing EJB JNDI is 1 way other way is
@EJB EJBCLASSNAME ejbclassobject; //this auto initialize and create object.
Initializing like above is standard or it is an extended support from some app server.

Hi,
Follow my opinion:
1 - With JPA we have a domain classes. Now its better to use entity as manage-bean for JSF or manage bean should be saparated.
>> I think that Managed-Bean must be separated, because you can need to bind you Visual Components to it too.
2 - Using DTO (Data Transfer Object) is good practice or not in JEE5.
>> You can put your Entity as a member of you Managed Bean.
3 - Simplicity or Complexity but with EntityManager I feel no need of DAO but I am used to with DAO pattern. So again as best practice I have to make 1 session bean as DAO and call it from all the session bean where I write business logic or forget about DAO session bean and call EntityManager from all session bean everywhere.
>> For CRUD operation I don't create a additional class, but for complex business logic, you can use a separated class (Business Manager)
Best regards

Best practices for file I/O within producer/consumer loops

I'm looking to add file recording and playback functionality to a pre-existing data collection program. The original progam is based on a Moore-style state machine, which I have added four additional states to. They are: Record Start, Record Stop, Playback Start, and Playback Stop.
What I have done, and what has since been identified as poor programming practice, was to "initialize" (either create or load) the appropriate file within the state machine loop during the "Start" command (for record or playback functionality), and then provide the file reference as an indicator, which is linked to for the appropriate read or write operation(whether I'm playing back or recording). The actual I/O occurs within the the Consumer Loop. (screenshots attached).
This is my first labview project outside of tutorials or other small examples, so any advise and constructive criticisims are welcomed. Specifically with regards to file IO and refnum routing (it gets a little hairy in the consumer loop)!
I'm running Labview 8.6 on Vista business.
Attachments:
Playback Start.JPG ‏71 KB
Consumer_Loop.JPG ‏122 KB
File Playback subvi.JPG ‏14 KB

jamoore84 wrote:
Ben,
Thanks for the suggestion. I think it's a little outside my ken at this moment, but I'll look it over. Despite any grevious coding transgressions, I have experienced some limited success with the current setup. While the use of Action Engines/Functional Globals may constitute the best practice, I might revise my post to read "acceptable and/or easily absorbable practices" instead.
Are there any other opinions on this? Let me start by listing a problem and posing a question:
Problem: I am able to playback a file only once. Subsequent attempts at file playback do not work.
Thanks in advance,
jimmy
HI Jimmy,
I don't give up easy.
Let me try to exaplain the issue with race conditons with a contrived example, the
"Command by Mail box" case.
Imagine you had a job where you recieved your orders via an old fashioned mail box. You never really saw your supervisor but relied on getting orders via the mail box. Now imagine the mail box could only hold one message and any time a new message was inserted, the old one would fall into the trash.
So you come to work each day check your mail box do what was ordered and everyone is happy.
The next day you come in and without your knowing, you are assigned to do the work of two bosses. So as long as you check your mail box more often than the two of them assigne work everything is fine... until you take a day of vacation! So you come back in the day after vacation and check you mail box and work a way until you catch hell for ignoring orders !?! Wel it turn out the order from boss 1 was replaced by the order from boss 2 while you were on vacation. Oh bother!
How can we fix it?
1) Expand the mail box so it hold more than one order. You just process them in the oder they are recieved.
2) Change to mail box to not accept a new order until you have removed the old.
Now back to reality!
Local variables act like the funky mail box. The last message insterted over-rides the previous.
Multiple variable writer ar like multiple bosses.
Queues operate like an expanded mail box, letting you handle each message in order.
Action Engines operate like the "mutexed" mail box.
Why I don't want to encorage you to "just patch up" what you have.
All of the less than ideal solutions either over-sample (Check e-mail twice as often as bosses assign work, waste CPU, an exercise in futility if you are coding in a non-Real-Time envirionment like Windows) or use a mutex to control access to the shared resource ( in this case local varialbes).
LV offers mutexes through semaphores (found on the syncronization pallete) but...
WHY WORK SO HARD?
In my AE Nugget I explain that the exection of an AE is automatically protected by LV. So in the long run it will actually be easier to learn how to use the AE programming construct than it will be to learn how to solve the problem without them.
so GO FOR IT! Take the lazy route and learn how to use the AE construct. Use the Syncronization pallete for Queues.
Just trying to help,
Ben
Ben Rayner
I am currently active on.. MainStream Preppers
Rayner's Ridge is under construction

Best practice to create views

Hi,
I've a question about best practice to develop a large application with many complex views.
Typically at each time only one views is displayed. User can go from a view to another using a menu bar.
Every view is build with fxml, so my question is about how to create views and how switch from one to another.
Actually I load fxml every time the view is required:
FXMLLoader loader = new FXMLLoader();
InputStream in = MyController.class.getResourceAsStream("MyView.fxml");
loader.setBuilderFactory(new JavaFXBuilderFactory());
loader.setLocation(OptixController.class.getResource("MyView.fxml"));
BorderPane page;
try {
     page = (BorderPane) loader.load(in);
     } finally {
          if (in != null) {
               in.close();
// appController = loader.getController();
Scene scene = new Scene(page, MINIMUM_WINDOW_WIDTH, MINIMUM_WINDOW_HEIGHT);
scene.getStylesheets().add("it/myapp/Mycss.css");
stage.setScene(scene);
stage.sizeToScene();
stage.setScene(scene);
stage.sizeToScene();
stage.centerOnScreen();
stage.show();My questions:
1- is a good practice reload every time the fxml to design the view?
2- is a good practice create every time a new Scene or to have an unique scene in the app and every time clear all elements in it and set the new view?
3- the views should be keep in memory to avoid performace issue or it is a mistake? I think that every time a view should be destroy in order to free memory.
Thanks very much
Edited by: drenda81 on 21-mar-2013 10.41

>
>
My questions:
1- is a good practice reload every time the fxml to design the view?
2- is a good practice create every time a new Scene or to have an unique scene in the app and every time clear all elements in it and set the new view?
3- the views should be keep in memory to avoid performace issue or it is a mistake? I think that every time a view should be destroy in order to free memory.
In choosing between 1 and 3 above, I think either is fine. Loading the FXML on demand every time will be slightly slower, but assuming you are not doing something unusual such as loading over a network connection it won't be noticeable to the user. Loading all views at startup and keeping them in memory uses more memory, but again, it's unlikely to be an issue. I would choose whichever is easier to code (probably loading on demand).
In choosing between reusing a Scene or creating a new one each time, I would reuse the Scene. "Clearing all elements in it" only needs you to call scene.setRoot(...) and pass in the new view. Since the Scene has a mutable root property, you may as well make use of it and save the (small) overhead of instantiating a new Scene each time. You might consider exposing a currentView property somewhere (say, in your main controller, or model if you have a separate model class) and binding the Scene's root property to it. Something like:
public class MainController {
private final ObjectProperty<Parent> currentView ;
public MainController() {
    currentView = new SimpleObjectProperty<Parent>(this, "currentView");
public void initialize() {
    currentView.set(loadView("StartView.fxml"));
public ObjectProperty<Parent> currentViewProperty() {
    return currentView ;
// event handler to load View1:
@FXML
private void loadView1() {
    currentView.set(loadView("View1.fxml"));
// similarly for other views...
private Parent loadView(String fxmlFile) {
    try {
     Parent view = FXMLLoader.load(getClass().getResource(fxmlFile));
     return view ;
    } catch (...) { ... }
}Then your application can do this:
@Override
public void start(Stage primaryStage) {
   Scene scene = new Scene();
   FXMLLoader loader = new FXMLLoader(getClass().getResource("Main.fxml"));
   MainController controller = (MainController) loader.getController();
   scene.rootProperty().bind(controller.currentViewProperty());
   // set scene in stage, etc...
}This means your Controller doesn't need to know about the Scene, which maintains a nice decoupling.

Best practice for adding text to Flex container?

Hi,
I'm having some troubles to lay a TextFlow class out properly
inside a Flex container. What's the best practice to achieving
this, for example adding a lot of text to a small Panel?
Is it possible to pass anything other than a static width and
height to DisplayObjectContainerController constructor, or is this
not the place to implement this? I guess what I am looking for is
the layout logic I'd normally pack into a custom Flex component and
implement inside measure() and so on.
My use case: a chat application which adds multiple TextFlow
elements to a Flex container such as Panel. Or use TextFlow as a
substitute for UITextField.
Some example code would help me greatly.
I'm using Flex 3.2.
Regards,
Stefan

Thanks Brian, the example helps. However problems quickly
arise if I modify it slightly to this (please compile it to see):
<?xml version="1.0" encoding="utf-8"?>
<mx:Application xmlns:mx="
http://www.adobe.com/2006/mxml"
layout="absolute" initialize="init()">
<mx:Script>
<![CDATA[
import flashx.textLayout.compose.StandardFlowComposer;
import
flashx.textLayout.container.DisplayObjectContainerController;
import flashx.textLayout.container.IContainerController;
import flashx.textLayout.elements.TextFlow;
import flashx.textLayout.conversion.TextFilter;
private var _container:Sprite;
private var _textFlow:TextFlow;
private function init():void
_container = new Sprite();
textArea.rawChildren.addChild(_container);
var markup:String = "<TextFlow xmlns='
http://ns.adobe.com/textLayout/2008'><p><span>Hello
World! Hello World! Hello World! Hello World! Hello World! Hello
World! Hello World! Hello World! Hello World! Hello World! Hello
World! Hello World! </span></p></TextFlow>";
_textFlow = TextFilter.importToFlow(markup,
TextFilter.TEXT_LAYOUT_FORMAT);
_textFlow.flowComposer.addController(new
DisplayObjectContainerController(_container, 200, 50));
_textFlow.flowComposer.updateAllContainers();
]]>
</mx:Script>
<mx:Canvas width="100" height="100" id="textArea" x="44"
y="46" backgroundColor="#F5EAEA"/>
</mx:Application>
What is the best way to make my textflow behave like a
'normal' UIComponent in Flex? Should I use UIComponent instead of
Sprite as a Container? Will that take care of resize behaviour?
I have never before needed to use rawChildren.addChild for
example, maybe you can explain why that's needed here?
I realise that the new Textframework works on an AS basis and
is not Flex or Flash specific, but this also poses some challenges
for those of us using the Flex framework primarily.
I think it would help to have some more basic examples such
as using the new text features in a 'traditional' context. Say for
example a TextArea that is just that, a TextArea but with the
addition of inline images. I personally feel that the provided
examples largely try to teach me to run before I can walk.
Many thanks,
Stefan

Row level security with session variables, not a best practice?

Hello,
We are about to implement row level security in our BI project using OBIEE, and the solution we found most convenient for our requirement was to use session variables with initalization blocks.
The problem is that this method is listed as a "non best practice" in the Oracle documentation.
Alternative Security Administration Options - 11g Release 1 (11.1.1)
(This appendix describes alternative security administration options included for backward compatibility with upgraded systems and are not considered a best practice.)
Managing Session Variables
System session variables obtain their values from initialization blocks and are used to authenticate Oracle Business Intelligence users against external sources such as LDAP servers or database tables. Every active BI Server session generates session variables and initializes them. Each session variable instance can be initialized to a different value. For more information about how session variable and initialization blocks are used by Oracle Business Intelligence, see "Using Variables in the Oracle BI Repository" in Oracle Fusion Middleware Metadata Repository Builder's Guide for Oracle Business Intelligence Enterprise Edition.
How confusing... what is the best practice then?
Thank you for your help.
Joao Moreira

authenticating / authorizing part is take care by weblogic and then USER variable initialized and you may use it for any initblocks for security.
Init block for authenticating / authorizing and session variables are different, i guess you are mixing both.

Best Practice for calling CFC

Hi,
In a web application, if I need to call a CFC method from a different CFC, what would be considered as the best way of doing it?
For example, let's say I have two components: Customer and Product. From a method functionA in Customer, I would like to call functionB in Product. I can do one of the following, but which way is best practice and why?
1. Create a Product object in functionA, and use it to call functionB
<cfcomponent name="Customer">
     <cffunction name="functionA">
          <cfset productObj = createObject('component', 'Product')>
          <cfset productObj.functionB()>
     </cffunction>
</cfcomponent>
2. Pass a Product object when we initialize a Customer object, and use that to call functionB
<cfcomponent name="Customer">
     <cffunction name="init">
          <cfargument name="productObj">
          <cfset variables.productObj = arguments.productObj>
     </cffunction>
     <cffunction name="functionA">
          <cfset variables.productObj.functionB()>
     </cffunction>
</cfcomponent>
3. Assume that Customer object has access to the Product object in the application scope
<cfcomponent name="Customer">
     <cffunction name="functionA">
          <cfset application.productObj.functionB()>
     </cffunction>
</cfcomponent>
Thank you very much.

The first two are fine. If the CFC being called is always gonna be the exact same one, then there's no prob directly referencing it in the calling CFC. If the CFC could vary, then pass it in.
If you're only using the CFC transiently, then you could use <cfinvoke> as well in this case.
Directly accessing an application-scoped CFC within a method is poor practice.
Adam

Best practice for initializing objects in a JSF backing bean?

Hi,
What is the best practice for initializing some objects in the JSF to-page backing bean before the to-page is displayed for the first time? The initialization would vary and depend upon a command link in the from-page.
Regards,
Al Malin

f:view has two new attributes in 1.2: beforePhase and afterPhase
which allows you to specify a phase listener method
which will be called before and after the view is processed.

Slow starup of Java application - best practices for fine tuning JVM?

We are having problems with a java application, which takes a long time to startup.
In order to understand our question we better start with some background info. You will find the question(s) after that.
Background:
The setup is as follows:
In a client-server solution we have a win xp, fat client running java 1.6.0.18.
(Sun JRE). The fat client containt a lot of GUI, and connects to a server for DB access. Client machines are typical 1 to 3 years old (there are problems even on brand new machines). They have the client version of JRE - standard edition installed (Java SE 6 update 10 or better) Pretty much usual stuff so far.
We have done a lot of profiling on the client code, and yes we have found parts of our own Java code that needs improving. we are all over this. Server side seems ok with good response times. So far, we havent found anything about shaky net connections or endless loops in the java client code or similiar.
Still, things are not good. Starting the application takes a long time. too long.
There are many complicating factors, but here is what we think we have observed:
There is a problem with cold vs. varm starts of the application. Apparently, after a reboot of the client PC - things are really, really bad - and it takes (sometimes) up to 30-40 secs to start the application (until we arrive at the start GUI in our app).
If we run our application, close it down, and then restart
without rebooting, things are a lot better. It then usually takes
something like 15 - 20 sec. which is "acceptable". Not good, but acceptable,
Any ideas why?
I have googled it, and some links seems to suggest that the reason could be disk cache. Where vital jar are already in disk cache on th warm start? Does that make any sense? Virus scanners presumable runs in both cases.
People still think that 15 - 20 sec in start up on the warm start is an awful long time, even though there is a lot, a lot, of functionality in the application.
We got a suggestion to use IBMs JRE - as it can do some tricks (not sure what) our SUN JRE cant do concerning the warm and cold start problem. But thats is not an option for us. And noone has come up with any really good suggestions with the SUN JRE so far?
On the Java Quick Starter (JQS) -
improves initial startup time for most java applets and applications.
Which might be helpful? People on the internet seem more interested
in uninstalling the thing than actually installing it though?
And it seems very proprietary, where we cant give our Jar files to it?
We could obviously try to "hide" the problem in some way and make it "seem" quicker. Where perceived performance can be just as good as actual performance. But it does seem a bad solution. So for the cold start we will probably try reading the jar files and thereby have them in disk cache before startup of our application. And see if that helps us.
Still, ok the cold start is the real killer, but warm start isn't exactly wonderfull either.
People have suggested that we read more on the JVM and performance.
java.sun.com.javase/technologies/performance.jsp
java.sun.com.docs/hotspot/gc5.0/gc_tuning_5.html
the use of JVM flags "-Xms" "-Xmx" etc etc.
And here comes the question .. da da ...
Concerning various suggested reading material.
it is very much appreciated - but we will like to ask people here - if it is possibe to get more specific pointers. to where the gold might be buried.
I.e. in a an ideal world we would have time to read and understand all of these documents in depth. However, in this less than ideal world we are also doing a lot of very timeconsuming profiling in our own java code.
E.g. java garbage collection is is a huge subject - and JVm settings also. Sure, in the end we will probably have to do this all very thoroughly. But for now we are hoping for some heuristics on what other people are doing when facing a problem like ours..?
Young generation, large memory pages, garbage collection threads ect. all sounds interesting - but what would you start with?
If you don't have info to decide - what kind of profiling would you be running and then adjust what JVM setting in your trials?
In this pressed for time scenario. Ignorance is not bliss. But makes it hard to pinpoint the or those JVM parameters to adjust. So some good pointers from experienced JVM "configurators" will be much appreciated!
Actually, If we can establish that finetuning of these parameters is a good idea, it will certainly also be much easier to allocate the time for doing so. - reading, experimenting etc. in our project.
So, All in all , what kinds of performance improvements can we hope for? 5 out of 20 secs on the warm start? Or is it 10 % nitpicking? Whats the ball park figure for what we can hope to achieve here given our setup? What do you think based on above?
Maybe someone out there have done some finetuning of JVM parameters in a similiar PC environments like, with similiar fat clients...? Finetuning so and so - gave 5 secs. So start your work with these one-two parameters?
Something like that - some best practices? Thats what we are hoping for.
best wishes
-Simon

Thanks for helpful answer from both you and kajbj.
The app doesn't use shared network drives.
What are you doing between main starts to get executed and the UI is
displayed?
Basicly, Calculating what to show in the UI. Accessing server - not so much, there are some reads from a cache, but the profiling doesnt indicate that it should be a problem. Sure, I could shift the startup time to some other slot, but sofar I havent found a place where the end-user wouldnt be annoyed.> Caching of something would seem most obvious. Normal VM stuff >seems unlikely. With profiling i basicly find that ''everything'' takes a lot longer in the cold start scenario. Some of our local java methods are going to be rewritten following our review. But what else can be tuned?You guys dont think the Java Quick Start approach, with more jars in disk cache will give something? And how should that be done/ what does people do?I.e. For the class loader I read something about
1.Bootstrap class loader
2.Extensions class loader
3.System class loader
and is wondering if this has something to do with the cold start problem?
The extensions class loader loads the code in the extensions directories (<JAVA_HOME>/lib/ext
So, we should move app classes to ext? Put them in one jar file? (We have many). Best practice about that?
Otherwise it seems to me that it must be about finetuning the JVM?
I imagine that it is a question about:
1. the right heap size
2. the right garbage collection scheme
Googling heap size for XP
CHE22 writes:
You are right; -Xms1600M works well, but -Xms1700M bombs
Thats one best practice or what?
On garbage collection, there are numerous posts, and much "masters of Java black art" IMHO, And according to profiling GC is not really that much of a problem anyway? Still,
Based on my description I was hoping for a short reply like "try setting these two parameters on your xp box, it worked for me" ...or something like that. With no takers on that one, I fear people are saying that there is nothing to be gained there?
we read:
[ -Xmx3800m -Xms3800m
Configures a large Java heap to take advantage of the large memory system.
-Xmn2g
Configures a large heap for the young generation (which can be collected in parallel), again taking advantage of the large memory system. It helps prevent short lived objects from being prematurely promoted to the old generation, where garbage collection is more expensive.
Unless you have problems with pauses, try granting as much memory as possible to the virtual machine. The default size (64MB) is often too small.
Setting -Xms and -Xmx to the same value increases predictability by removing the most important sizing decision from the virtual machine. On the other hand, the virtual machine can't compensate if you make a poor choice.
The -XX:+AggressiveHeap+ option inspects the machine resources (size of memory and number of processors) and attempts to set various parameters to be optimal for long-running, memory allocation-intensive jobs]
So is Setting -Xms and -Xmx and -XX:AggressiveHeap
best practice? What kind of performance improvement should we expect?
Concerning JIT:
I read this one
[the impact of the JIT compiler is obvious on the graph: at startup the time taken is around 500us for the first few values, then quickly drops to 130us, before falling again to 70us, where it stays for 30 minutes,
for this specific issue, I greatly improved my performances by configuring another VM argument: I set -XX:CompileThreshold=50]
The size of the cache can be changed with
-Xmaxjitcodesize
This sounds like you should do something with JIT args, but reading
// We disable the JIT during toolkit initialization. This
// tends to touch lots of classes that aren't needed again
// later and therefore JITing is counter-productiive.
java.lang.Compiler.disable();
However, finding
the sweet spots for compilation thresholds has been tricky, so we're
still experimenting with the recompilation policy. Work on it
continues.
sounds like there is no such straigth forward path, it all depends...
Ok, its good, when
[Small methods that can be more easily analyzed, optimized, and inlined where necessary (and not inlined where not necessary). Clearly delineated uses of data so that usage patterns and lifetimes are apparent. ]
but when I read this:
[The virtual machine is responsible for byte code execution, storage allocation, thread synchronization, etc. Running with the virtual machine are native code libraries that handle input and output through the operating system, especially graphics operations through the window system. Programs that spend significant portions of their time in those native code libraries will not see their performance on HotSpot improved as much as programs that spend most of their time executing byte codes.]
I have the feeling that we might not able to improve performance that way?
Any comments?
otherwise i was wondering about
-XX:CompileThreshold=50 -Xmaxjitcodesize (large, how large?)
Somehow, we still feel that someone out there should have experienced similiar problems? But obviously there is no guarantee that the someone should surf by here!
In c++ we used to just write everything ourselves. Here it does seem to be a question about the right use of other peoples stuff?
Where you are kind of hoping for a shortcut, so you dont have to read endless number of documents, but can find a short document that actually addresses your problem ... well.
-Simon
Edited by: simoncpm on Mar 15, 2010 3:43 PM
Edited by: simoncpm on Mar 15, 2010 3:53 PM

Best practice to initialize my instruments

Similar Messages

Maybe you are looking for