Solved

Performance issues with object remoting (MS.NET)

Posted on 2002-05-22
6
199 Views
Last Modified: 2010-04-02
We are in the process of developing an enterprise project management system using MS.NET (C#).

The internal structure of the system is based on object remoting, which provides a whole lot of great features, but also seems to carry with it a few intricate problems.

The problem is basically that when we access properties and methods of remoted objects, the bad performance forces us to use "ugly" solutions. Let me illustrate this with an example:

Let's assume that "PersonArray" holds 1000 remoted objects of the class "Person". Executing the following code then takes ages!

  foreach(Person CurrentPerson in PersonArrray)
  {
    CurrentPerson.Age++;
  }

If the corresponding code is executed on the server side, the code takes virually no time at all to execute. We have only found one solution that manages to speed the operation up, at least a bit, and that is to send the array to the server side and execute the loop there.

So, the problem is basically that looping over large collections of remoted objects takes ages.

Does anyone know how we can solve the problem? Any tips where we can find in-depth reading about object remoting?



Best regards
Karl Hagstrom
Bite Solutions AB
Sweden
0
Comment
Question by:Kalleman
[X]
Welcome to Experts Exchange

Add your voice to the tech community where 5M+ people just like you are talking about what matters.

  • Help others & share knowledge
  • Earn cash & points
  • Learn & ask questions
6 Comments
 
LVL 6

Expert Comment

by:snoegler
ID: 7027149
Hmm this is a typical task which should be DONE at server side. Imagine a SQL database:

UPDATE customers SET customer_id=customer_id+1

If the DB would act like your example, the client of the SQL Server would fetch all records (completely, like SELECT *), increase the field customer_id by one, and send each record back to the server. Now this would take ages :)

You should modify your design, handle processes like this completely on the server and provide a straight, orthogonal interface.
0
 

Author Comment

by:Kalleman
ID: 7027256
OK. I must say that I agree. Making trillions of calls to the server side, instead of making one bach call IS bad. However, my goal is to use an architecture that doesn't force me to implement trillions of methods on the server side like:

UpdatePhoneNumbersForPersons(...)
ChangeOwnerOfCar(...)
RemoveEmployeesFromCompany(...)
...
...

Being able to handle my objects directly on the client side isolates the client from the server and makes the code structure better. I have been trying (without satisfactory results) to implement a solution with disconnected objects, sort of a "check out" of business objects to talk in source code control terms.
0
 
LVL 6

Accepted Solution

by:
snoegler earned 150 total points
ID: 7027289
I have attended an MS event where the contributor pointed out that exactly what you're experiencing is what he expects to make problems with .NET remoting in "real life" programming.

You must understand that remoting is REALLY expensive and is (from my POV) much too comfortable. Consider what the architects of SQL did: they invented a low-level, orthogonal yet complete and manageable language for handling most tasks directly on the server, almost 20 year ago now.

In your case, i'd suggest an interface based on XML. On client side, you implement a light-weight API to help constructing the XML for server side invokes. Sample:

<server-invoke>
 <target what="customer" refid="82">
  <alter what="phone">
   <destval>(+43)738-03939</destval>
  </alter>
 </target>
</server-invoke>

or, in the case of your age example:
<server-invoke>
 <target what="customers" select="{custom SQL WHERE clause maybe}">
  <alter what="age">
   <destval>age=(age+1)</destval>
  </alter>
 </target>
</server-invoke>

In the latter example, the statement 'age=(age+1)' could be passed directly as an argument to an SQL UPDATE statement. If you manage to construct an API like this, you'll experience a performance about 1E+5 times faster than using remoting for that.
0
Industry Leaders: We Want Your Opinion!

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

 

Expert Comment

by:hmatchen
ID: 7028516
A sneaky possibility is to build a stored procedure dynamically, post it to the server, execute it, then delete it.  This has the advantage of flexibility, but requires that the user's workstation have that level of access to the db server.  Update queries don't have result sets, so there's no debris.
0
 
LVL 49

Expert Comment

by:DanRollins
ID: 7029117
I think you should just code the PersonArray remote object with functions that iterate across all of its Person objects, or a specified subset of them:

    PersonArray.UpdateAll( AGE, +1 );

    PersonArray.Update( AGE, +1, "lastname > 'S'" );

The idea being that the Person object never gets moved across the network.

-- Dan
0
 

Author Comment

by:Kalleman
ID: 7058852
Thanks for the feedbak, even if it means I probably have to completely redesign the data tier of our system. *doh*

0

Featured Post

Free Tool: IP Lookup

Get more info about an IP address or domain name, such as organization, abuse contacts and geolocation.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Introduction This article is the first in a series of articles about the C/C++ Visual Studio Express debugger.  It provides a quick start guide in using the debugger. Part 2 focuses on additional topics in breakpoints.  Lastly, Part 3 focuses on th…
Container Orchestration platforms empower organizations to scale their apps at an exceptional rate. This is the reason numerous innovation-driven companies are moving apps to an appropriated datacenter wide platform that empowers them to scale at a …
The goal of the video will be to teach the user the difference and consequence of passing data by value vs passing data by reference in C++. An example of passing data by value as well as an example of passing data by reference will be be given. Bot…
The viewer will be introduced to the member functions push_back and pop_back of the vector class. The video will teach the difference between the two as well as how to use each one along with its functionality.

707 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question