• February 27, 2025, 07:49:25 PM
  • Welcome, Guest
Please login or register.

Login with username, password and session length
Advanced search  

News:

This Forum Beta is ONLY for registered owners of D-Link products in the USA for which we have created boards at this time.

Author Topic: VPN using Client/Server apps  (Read 3016 times)

kramerind

  • Level 1 Member
  • *
  • Posts: 2
VPN using Client/Server apps
« on: December 22, 2010, 10:23:06 AM »

I have a client with a small office who has a client/server application they use regularly in the office.  It's stored on a DNS-323 and with 7 people in the office banging on it, it performs just fine.  Some office members are able VPN from home, when they try to use the application it's extremely slow.   Any suggestions?

The network is simple....

  Internet --- Cable Modem --- SonicWall Firewall --- DNS-323

Other than clients machines being on the inside of the firewall and network printers and a usb printer to the DNS-323, that's it.   No file server or anything else.  The SonicWall is handing out IP addresses.  The DNS-323 has a static IP.

I've tried various settings on the SonicWall and on the client PC (MTU size, etc).  No improvements.  There is good bandwidth and speeds between the remote user and local speeds and bandwidth are good.  With only one user, the VPN user coming in remotely, on the network, the speed is very slow in the client/server app.   The drive lights on the DNS box are going like crazy when the remote user navigates through the application remotely. 

Internally, with seven users beating on the DNS box running the same application, response times are good.   I know VPN is creating some overhead, but are there any suggestions?   Am I going to be forced to put a box on the network for the sole purposes of running terminal server or some other remote desktop application in lieu of VPN?

Any suggestions would be appreciated.
Logged

fordem

  • Level 10 Member
  • *****
  • Posts: 2168
Re: VPN using Client/Server apps
« Reply #1 on: December 22, 2010, 12:34:23 PM »

First - a question - you say a client/server application stored on a DNS-323.  Are you sure that you are, in fact, running a client/server application?

For this to happen, you would have had to hack the DNS-323 to get the server portion of the application running on it - you're going to need to install some sort of database server along with that application and given the limited memory and processing power of the DNS-323 - I'd say that would be quite a feat.

In a true client/server server scenario, the application is split into two portions - a client side, that runs on the client, and a server side, that runs on the server - the idea behind this is to reduce network traffic.

A relational database management system running on a computer and manipulating a database stored on a different computer will frequently generate significant volumes of network traffic - just as an example - a request to display all records with a particular description (show me all the 1TB drives in inventory), would require the workstation to search the entire database by requesting every record and examining the description field.

If this same system were designed as a client/server application - the client side application would send an SQL (structured query language) query to the server side application, which would then do the search, and return just the dozen or so records that match the search criteria.

A true client/server application will run using remote access with very little degradation, but the degradation you describe suggests that your application is not client/server but merely a multiple user RDBMS with a shared database.
« Last Edit: December 22, 2010, 12:37:07 PM by fordem »
Logged
RAID1 is for disk redundancy - NOT data backup - don't confuse the two.