Revlet deployment - a network issue?

[frullone]

Hi,

we have developed a revlet intranet application which provides navigation between different cards, access to resources with http and ftp protocols.The application works perfectly in
our LAN, but when the application is launched on the customer's LAN, performances are extremely low for two minutes, after this time the performances are normal.
A way to workaround this enormous delay is to remove client's configuration of the gateway. (!!??)

Installing a basic html website, instead of the revlet, generates a normal behaviour.
Removing ftp access to resources does not change the situation.
The customer uses CISCO switches and network devices, and the configuration is simply:
Client IP 10.68.22.117 /24
Server IP 10.68.22.224 /24
DNS IP 10.68.22.223 /24
Gateway IP 10.68.22.254 /24

got any ideas??

[Mark]

Hi frullone,

What exactly is slow? You click a button in the revlet and then it is supposed to do something, but this happens 2 minutes later, or what?

Best,

Mark

[frullone]

That's right, Mark,
suppose the revlet opens with a card with 5 buttons, and each button goes to a new card on its mouseUp. The expected behaviour is that the new card opens with a new image and some other buttons and so on, so that a tree-like structure can be navigated.

Nothing special, indeed. And as a matter of fact it works.... as soon as we use it on our LAN. What happens on our customer's network is that when we click on a button, the revlet immediately sends a couple of http "GET"s , but for 30 secs nothing happens (we've sniffed network traffic), after those 30 seconds the remaining "GET"s are sent, and then the destination card displayed. In those 30 secs we see NO network traffic! The network is not loaded and nothing strange happens, other than this blackouts.
When we've reached the "leaf" of the tree, about 150 secs have passed since the openstack, and navigation becomes normal, transitions between cards are immediate.

How have we patched this problem, by now? Removing the gateway address from the client's IP settings. When a gateway is specified, blackouts happen again.
On the other hand, when we test it "in house", nothing like that can be observed. What could be the cause of this enormous delays? Thanks for your help!!

[Mark]

Hi frullone,

Tell your client to use a different gateway, not ending in 254. 254 is a broadcasting address. Try 253 This means a reconfiguration of the network, but it should make a difference, particularly on Windows machines.

Let me know if this helps.

Kind regards,

Mark

[frullone]

Dear Mark,
thank you for your help , but the subnet broadcasting address is 255, not 254. I'll try it, but I don't think that it will work anyway. Any other idea?

[Mark]

Hi Frullone,

Yes, you're right. Actually, I checked my set-up and my gateway is at *.254 too! This works fine for me.

Can you tell a little more about how you load the media? Do you load pictures by script or do you have filenames set to web addresses? Can you post your scripts? How many pictures and other media are we talking about? Do you have the alwaysBuffer of all those objects et to true or false?

Mark