Re: RDP Connections - Freeze or Reconnect

Thanks TP for your advice, much appreciated.

I agree with you, I think this is network related. Just running another ping test as I forgot to set the -t parameter and walked away from the screen! It's been a long hot day!

Yes the Red X was against the 1GB LAN connection icon. The cable comes from the switch and then connects into what I can only describe as an extension cable with a female/female connector. I immediately swapped the connector and cable out and plugged the cable coming out of the switch into the spare 100mb NIC adapter but still using the connector and extra cable. The old extension cable is still connected. Everything else has been replaced with equipment which was not in use.

The connection of the 100mb LAN from then on appeared much more stable. I didn't see any red x's appearing at any time. I also adjusted the autodetect setting from auto to 100mb Full Duplex on the spare 100mb adapter and disabled the 1GB adapter. I cannot figure out which switch port the cable is using because nothing is labeled! It is now however plugged into a different port but not sure where in relation to the 1GB switch port. I could have got them to check the port lights when I turned the server off. I might do that tomorrow. I don't know if I can manage the switch in anyway. Not sure if it is managed or what model/make it is or what the IP Address is. Another company provided the switch and cabling. This is the partner coming in tomorrow.

Unfortunately nothing is logged in the event viewer. I was hoping to find something about the adapter.

Still running ping test...

If I logon and test the RDP connection with a dialup modem through RRAS (Terminal server is also a RAS Server), would you expect it to misbehave if it was I switch problem? I am guessing my dial up connection doesn't go through the switch in anyway?


"TP" <tperson.knowspamn@xxxxxxxxxxxxxxx> wrote in message news:uPSZX4t1IHA.3884@xxxxxxxxxxxxxxxxxxxxxxx
Hi Sarah,

Everything you have described so far points to a networking issue, and has nothing to do with TS. I suspect if you have your users do a continuous ping the problems will be apparent.

Some key points:

- Having Exchange on the same server will *not* cause frequent disconnects/reconnects. The other Partner is dead wrong on this point.

- You mentioned seeing a red X on the server's network connection in an earlier post. I assume you are referring to the NIC connection object on the server (if I understood you correctly). This is another sign of a network issue and is unrelated to TS, Exchange, etc.

- Have you tried plugging the server into a different port on the switch, preferrably one that is several ports away from the current one? I say several ports away because in some cases adjacent switch ports will share the same [faulty] part.

- Are there any messages in the server's or workstation's logs related to networking? Like entries from the adaptor saying that it has been disconnected, followed by a message saying it has been reconnected, or from dnsapi, dhcp, tcpip saying it has resumed normal operation, etc?

- Have you tried manually setting the speed of the server's NIC, for example, backing it down from Gigabit to 100Mbps? Is Spanning Tree disabled on the switch?

- You mentioned replacing the server's network cable, what about the patch cable leading from the punchdown block to the switch (if it exists--disregard if the server connects directly to the switch)?
- If there are any network jacks between the server's NIC and the switch port, try using a long network cable in order to bypass them and connect directly to the switch. Sometimes network jacks can be pulled a bit and then the signal will be poor.

From everything you have said my guess would be the switch is having a problem. This is based on you updating drivers, changing out cables, switching to a different NIC, etc., and your statement that the switch is the only thing left unchanged.

Thanks.

-TP

Sarah Kingswell wrote:
Thanks for all your advice, I do appreciate your help.

The customer has another Partner coming in tomorrow to look at the
issue and they believe the problem is because the server is
overloaded as it also runs Exchange. It has been running exchange
for 5 years on the same box and never has it had a problem, so I
completely disagree with them! I am really keen to find an answer
today. Lan users connect directly to internal IP of server through RDP.
Remote users connect with VPN and then the RDP.
I have also connected directly with the server with a port forward on
3389. All instances cause problems.

I asked them to find the old switch that was upgraded a few years ago
to test but typically they can't find it. I can logon to the server
with a modem and then RDP to see if this makes any difference. Not
sure what this will prove. Seriously think something is corrupted or
there is a hardware issue. Testing the switch will be the best idea.
I just hope the company going in tomorrow can do this without having
to sell the customer a new system!

"moncho" <moncho@xxxxxxxxxxxxxxxxx> wrote in message
news:q5t8k.6564$cW3.4198@xxxxxxxxxxxxxxxxxxxxxxx
Sarah Kingswell wrote:
The client has disconnected the router from the network and internal
users accessing the TS over RDP are still reconnecting frequently.
Nothing logged in event viewer for both server and on client end.

If I look at the licence manager for TS I see

Windows 2003 - Terminal Serer Per User CAL
Type=Open
Total=20
Available=20
Issued=0

Not sure if this should be showing a number of issued licences. I
have re-actived the licences and the error has not been reported in
the event log since doing this. Does this look right to you?

Every user accessing the terminal server via RDP is seeing the
session reconnect. It reconnects in seconds and the screen is back
to where it was when it left it. Nothing is logged anywhere
suggesting a problem. I am going to run a ping -t to see if any
packets are being lost over the LAN. May imply it is an issue with
the switch as everything else has been replaced.

Any ideas where I go next?

You mentioned that the server is at a remote site from where you are
at. Is there anyone who can watch the power lights on the switch to see
if anything weird is going on?

Are the remote users connecting via a VPN or does the router port
forward to the TS server? Also what type of router is it?

I have had similar disconnect problems with ADSL and Cisco SSL-VPN
but the reconnects only happen 2 or 3 times a day or less.

Based on what you stated above, the internal users are now having
issue too, correct?

Do you have an extra switch that you plug into the router and do
some testing without messing with the rest of the users?

I am wondering if you need a new UPS for the switch, just to make
sure it is not loosing power or if the power is fluctuating.

Do you have the ability to do an RDP session to a local machine
from a remote site, not a VNC connection?

As for your User Licensing situation, Windows does not keep track of
these (not until W2K8). What you have looks fine.

moncho


"Sarah Kingswell" <sarahkingswell@xxxxxxxxxxx> wrote in message
news:%23tNLHds1IHA.3860@xxxxxxxxxxxxxxxxxxxxxxx
There is nothing logged in the event viewer about this. A few odd
warnings but nothing that I haven't seen before.

I will run another windows update tonight and I have just asked the
customer to turn off the ADSL router to see if this improves the
connection with the internal users.

Yesterday we put a new APC-UPS on the server and installed the
latest version of powerchute. I have since removed UPS and
software thinking this was the problem. What did happen yesterday
which may be related is that the server lost power from the UPS
because the UPS connections on the back are really poor and the
power cable just fell out. This mean't the server shutdown
without warning. I have reset winsock thinking this might be
corrupted. Also now using a second NIC but with no improvements.
Could it be the switch? One thing that I have noticed is that if I go directly to the
server console and run RDP to 127.0.0.1 - it is stable.

"moncho" <moncho@xxxxxxxxxxxxxxxxx> wrote in message
news:5Mq8k.14504$Ri.10985@xxxxxxxxxxxxxxxxxxxxxxx
Sarah Kingswell wrote:
This problem is getting more bizzare by the day and is now
driving me crazy!!!

I thought this problem was resolved but today it would appear it
has got much worse. I went out to the site today and all was
well until all of the internal terminal users started getting a
'reconnecting to session' windows when they were logged onto the
terminal server. It was reconnecting very frequently (every 2
minutes or so). Internal Users have never experienced this in
all the 5 years that they have been running the server/ I left
site thinking I had fixed the issue by replacing the Network
cable. I noticed that the network connection had a red x on the
server and would then reconnect intermittently. To be sure I
plugged a new cable into a different ethernet port on the
server. This seemed to work. I got home and logged onto the terminal server with my Vista
laptop. The RDP session keeps reconnecting but it drops the
connection at 3 seconds past the minute everytime. 10:13:03,
10:14:03, 10:15:03 etc. Thinking this was an issue with Vista, I
tried RDP connection from my other XP laptop. Same type of
thing happens.

This is an interesting issue. Please post follow ups for future
reference.

Is there anything in any of the Event Logs?


I then used my Vista laptop from home to logon to another PC
(lets say computer Z - Windows XP) using VNC which is in a
remote location. From computer Z I logged into the Terminal
Server with RDP. Therefore connecting to the terminal server
with a different internet connection. This was stable and didn't
reconnect. I am not sure what this means exactly but could it rule out the
theory that the ADSL line / Router in the Terminal Server
building is the problem?

This could rule out the ADSL and router but VNC may be
compensating for the bad connection where as RDP is just killing
it immediately.

Today I also notice that the TS is issuing temporary licences.
The licences were installed and activated 5 years ago. I will
need to fix this but I don't know why this has all of a sudden
stopped working.

Do you use User or Device Licenses?

What is the status of your license server? Anything unusual
there?

If anyone can save me from going mad... please HELP!!

I am trying to stick with hardware because that is the easiest
place to look (IMO).

Do you have an extra ADSL router that you could test with?

Are there multiple NIC's in the server? If so could you use the
second one if it is not in use already?

Has the server been updated with the latest Windows Updates and
NIC drivers?

Did this start when some software or hardware was
updated/upgraded? moncho


"moncho" <moncho@xxxxxxxxxxxxxxxxx> wrote in message
news:BCP7k.5301$L_.830@xxxxxxxxxxxxxxxxxxxxxxx
Sarah Kingswell wrote:
I am having a really frustrating issue with RDP connecting to a
Windows 2003 SP2 Server. For several years the Terminal
Server has been working well but just recently remote users
have been complaining that the RDP session frequently lock up
or reconnects with a big red x in the top right hand corner of
the screen. It is happening so frequently that accessing the
terminal server remotely is becoming unusable. I have seen
this before and it happens very occassionally. LAN users
access the terminal server with RDP without issue. This makes
me believe it has something to do with the Router or ADSL Line.

I have upgraded the router firmware to the latest version
Applied all of the latest updates to windows
Updated network drivers etc
Rebooted server
Changed MTU settings

Nothing seems to be making any improvements.

Does anyone have any ideas how I can troubleshoot this further.
Workstations are a mixture of XP sp2 and Vista Sp1

Does the Router from the ADSL line go into the same network
switch that the LAN users are on?

If so and the LAN users are fine, I would suspect the ADSL line.

moncho

.

Relevant Pages

  • Re: Outgoing POP3 email missing/lost/not received
    ... Funny thing is that I have had this ISP for 8 years and it has always been ... It looks like when you last ran CEICW, you set the ISP's mail server to: ... Internet Connection Wizard. ... After the wizard completes, the following network connection ...
    (microsoft.public.windows.server.sbs)
  • Re: Outgoing POP3 email missing/lost/not received
    ... ISP's mail server instead of the domain name on the ... SUMMARY OF SETTINGS FOR CONFIGURE E-MAIL AND INTERNET ... Internet Connection Wizard. ... After the wizard completes, the following network connection ...
    (microsoft.public.windows.server.sbs)
  • RE: Problems with Permissions
    ... And SBS server is only take ... the role of an internal server. ... they are all configured to connected to internal network. ... g. Run the Configure Email and Internet Connection Wizard on SBS server. ...
    (microsoft.public.windows.server.sbs)
  • Re: Connection from remote computer to network SQL Server
    ... There is no firewall on the W2K machine acting as the SQL server. ... I tried making the SQL machine a "trusted" on the router. ... connection works. ... To find the IP address of your computer inside the network, ...
    (microsoft.public.access.adp.sqlserver)
  • Re: Networking Question - VLANs on SBS 2003 Premium SP1
    ... be sure you do not enable any DHCP server in internal network. ... You do not get any issue when you connect the SBS to the old router, ... On the Connection Type page, click Broadband, and then click Next. ...
    (microsoft.public.windows.server.sbs)