what can be causing the connection to drop??

Where to begin...?

Assuming that noone in Europe has latency/performance issues with the application in question, it would appear you're looking at a network-level issue.

In your question title, you say the connection drops.  Does it actually drop?  Do you have any client and/or server-side error/event logs you can provide to show this drop?

Or are you saying "drop" to describe the lag?  

Am I reading correctly that you have a bottleneck on data coming from your application server in Europe to client systems in the US (US -> EU 8Mbps; EU -> US 6Mbps)?

When the application hang occurs at the US site, does it impact all users at roughly the same time?  Or is it only a subset of users that experience the problem?

Need a bit more info to figure out where to dig in.

Application-specific and/or client server OS event log errors and/or warnings around the time of the lag would be a good start.

Also, what kind of routers?  Are they under your administrative control (you mentioned adding static routes)?

At a very fundamental level, you could start a

PING -T [IP OF EU HOST] > pingtest.txt 

Select allOpen in new window

on two or three client computers at the US site, let it run for a while (ideally until the next lag/drop event occurs), then check the results for latency or loss of connectivity.

Well, my one user says that its very slow sometimes and they said once it dropped.

I dont have access to the server that is hosting the application. So I cant see the logs for the server.

Im not sure if its a bottleneck or something with the local pc's... im tossing around in my head what can be causing it.

I cant say if its both users. I asked for a test accoung myself so that I can check to see if I experience the same issue when the one specific user who is on it frequently using it says they experience it.

would there be any logs defining this on my pc even though this is an MPLS connection to the server overseas?

My firewall/router is a sonicwall . I added the static route to the MPLS router which is a Cisco router that connects to the Cisco router overseas that was provided by our ISP vendor. I only have administrative control to my Sonicwall router.

When I run that ping test..what signifies in the log a drop or latency?? I know very little about  how to read those logs so sorry for the ignorance of the question.

where is that text file saved for the ping test command?

Well, my one user says that its very slow sometimes and they said once it dropped.

You should gather more detail from the user.  How often is "sometimes?"  How do they know it "dropped?"  What errors, if any, were reported by the application?

would there be any logs defining this on my pc even though this is an MPLS connection to the server overseas?

Is this a web app?  Or is there a client of some kind installed on the US computers?  If the later, then there *may*be application logs on the US computers.  Really depends on the app.  Who supports the app?

My firewall/router is a sonicwall . I added the static route to the MPLS router...

Is the MPLS router inside the firewall, or outside?  You might consider adding the static route to your DHCP configuration (assuming that's how the US client computers get their IP addresses).

When I run that ping test..what signifies in the log a drop or latency??

Here are the results from PING

C:\>ping -t 8.8.8.8

Pinging 8.8.8.8 with 32 bytes of data:
Reply from 8.8.8.8: bytes=32 time=66ms TTL=43
Reply from 8.8.8.8: bytes=32 time=52ms TTL=43
Reply from 8.8.8.8: bytes=32 time=56ms TTL=43
Reply from 8.8.8.8: bytes=32 time=53ms TTL=43
Reply from 8.8.8.8: bytes=32 time=52ms TTL=43

Select allOpen in new window

This will go on and on until you terminate with CTRL+C.  When you do, you will see

Ping statistics for 8.8.8.8:
    Packets: Sent = 5, Received = 5, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
    Minimum = 52ms, Maximum = 66ms, Average = 55ms

Select allOpen in new window

Look at "Lost = 0 (0% loss)" <- any number other than 0 indicates some loss.
As for latency, you really need a baseline when things are running as expected.  The Min/Max/Average numbers are what we're looking at - but you don't know what's abnormal until you know what's normal...

where is that text file saved for the ping test command?

Whatever directory you're in when you run PING.

Sorry - I should've said (?) that you need to open a command prompt first.

Start -> Run -> "cmd" -> ENTER

Your default path will be shown.  Typically C:\Users\You on recent Windows clients...

This is a web app? It is run overseas and maintained by the IT dept over there.

The MPLS router is outside. Its provided to us by Verizon.

How do I add a static route to DHCP? Its a windows 2008 server. Why is this better than setting it up in the router?

Here are some suggestions you could look into

1. Check your round trip time between both sites across the MPLS core


2. Check for MTU issues across the MPLS Core  

ping <remote_site> -l <size>

work your way downwards to identify you max. MTU

3. Capture a conversation between two hosts to see unusual things like lots re-transmissions etc

nazky,

I know very little about routing and MTU recommended size.

Can you elaborate on specifics that I should be looking for?

Right now, if I ping the remote site the time is 95ms. What should I do to check the round time and what is a good range I should be looking for?

What should the MTU size be?? when I use the ping command what should the <size> be?

How do I capture a conversatiion between the hosts?

The round trip time of 95ms is OK

To Check for MTU issues across the MPLS Core  say 10.1.1.1 is the remote host

run these command, the goal is to start from a low value and work up to the maximum allowable MTU of 1500.


ping 10.1.1.1 -l 1200

ping 10.1.1.1 -l 1300

ping 10.1.1.1 -l 1400

ping 10.1.1.1 -l 1500

If 1400 is successful but 1500 fail; work upwards from 1400 in smaller increments ... 20 maybe

Once you identify the last successful value that is your Maximum Transmission Unit over the MPLS core. Your host should not be sending data units larger than this value.

ok thank you I will try this..

one last question.. how do you see what the size of the data units you are sending?