Remote Desktop connection to server problem

[randle]

Hi

 

Recently I have been unable to open an RDP session to one of my servers. When attempting to connect, the Remote Desktop Connection window appears briefly then just disappears. A reboot of the server usually fixes this but it's no time before this becomes a problem again.

 

No firewall is activated in this instance and cannot connect to this server from any client/server. All connections to other servers/clients work fine so is definitely isolated to this one server.

 

I've also reset the RDP-Tcp (Listener) protocol on the server

 

Any ideas?

[SW-ICT]

make sure the server has the currect staic IP, Dns and everything... Then try the Start > run > mstsc /console

 

Let me know how you get on

[randle]

Static IP and DNS records are all in order. The server is functioning as normal on the network other than this so don't think it's anything quite so obvious unfortunately.

[SW-ICT]

has it got the same ip address to what is on the DHCP server... thats messed me around a little bit before

[SW-ICT]

Remove the network drive and reinstall again And thats all i can really think of at the top of my head sorry buddy

[randle]

DHCP only shows and deals with leased addresses, not static addresses so won't be a factor here. Thanks anyway.

[Admiral208]

anything in the event viewer that might shed any light??

[SW-ICT]

but it should say what address it should have..

[randle]

Nothing that seems relevant.

[randle]

but it should say what address it should have..

 

I have an exclusion range set up and this static IP address falls in this range so it won't show up in the address leases window like DHCP enabled clients would.

[SW-ICT]

with us it was something really silly. (not saying your silly) i think the DNS settings was wrong.. try checking them and so a cmd > "dns /flush" or its flush /dns

[randle]

If DNS settings for this server were wrong I'd be seeing other more important problems appearing like users not being able to access shared areas on this server etc. As mentioned this isn't just happening from the one computer so isn't a problem with the local DNS cache either.

 

Regardless of that DNS records for this server are present as expected.

[maniac]

What happens if you log onto the actual server while this is happening? I used to get issues like this ocassionally, and when I then logged onto the actual server it used to generate me errors saying the system was out of resources and was un-able to log me on or, or was logging me on with a local profile - all sorts of weird and wonderful messages like this. We also got issues with users accessing their home areas etc. at the same time. Used to happen ever 2 weeks or so to our storage server.

 

I did find a fix for it, but unfortunitely I can't actually remember what it was now. I know it entailed changing some registry values however.

 

Mike.

[randle]

Nothing untoward or helpful appears unfortunately. I am getting paged pool errors in the event log but don't know how relevant that is at the moment without further investigation.

[maniac]

Nothing untoward or helpful appears unfortunately. I am getting paged pool errors in the event log but don't know how relevant that is at the moment without further investigation.

 

Ah yes that sounds familiar, I'd have a guess that the two are linked.

[randle]

Interesting you think that. I'll take a deeper look and let you know

[m25man]

Sounds like you have another application/service running that is overlapping the RDP stack.

 

What do you see in a dos box when you run netstat -a

 

Is port 3389 bound and listening on the correct IP?

 

Try disabling any other non critical services or recently installed apps and restart the Terminal Services or reboot the server this may reveal which app is preventing your RDP from starting correctly.

[maniac]

It was event 2019 that was plaguing our error logs, I've just looked back at my notes. Is that the same error you're getting?

[randle]

Since rebooting I can log on so will be a little while before this starts being an issue again. When it does I'll check out what's been mentioned and report back.

 

I haven't been getting any errors in the event log but will double check next time.

 

Thanks for your replies

[RobBaxter]

It's something in windows update. *Fumes towards MS direction* if its Windows Server 2008 R2 (as i have) I am unsure as to if this happens on other versions.

 

Anywhoo... let me explain. When my servers installed SP1 for R2 back in feb, when they applied the update, which the server was "online" maybe services apeared in a state of wierd 1/2 up 1/2 down mode. RDP was one of them, while authentication seemed to be functioning, the moment it tryed to show me a desktop the application just terminated as if i closed it my self.

 

What I later found out, was while windows is applying updates "which in theory is only when it boots". Once the updates where applied, the RDP sessions worked fine.

 

Now, its very coincidental that a professional such as yourself seems to have the "same" symptoms. Try taking a look at windows update and see if its glitching out on you, as maybe this is a reason for the "wierd time to kick in" as we all know, windows update is scheduled, do they coincide?

 

Food for thought!

Edited June 20, 2011 by RobBaxter

[lactose]

In my case, it was after applying Windows Server 2008 SP2 - the console itself was waiting for a person to click "Ok" that SP2 had finished installing.

Sometimes, there's an additional whrily-gig of dos-window activity related to the Service Pack CAB file after you log in.

Funny thing: even if you RDP /console (or /admin), it won't let you in, you have to go THE ACTUAL console, login and click 'OK.'

Maybe because some piece of RDP itself was being updated.

After clickink 'ok' at console, all is well, RDP is happy, etc.

[ass17]

This is how I fix it:

 

It turns out that the cause for the problem is two important Windows security updates:

 

KB 2621440

KB 2667402

These updates were part of MS012-20, and were downloaded to the servers sometime in the second half of March 2012, if their settings were configured so.

 

The solution to the problem begins by uninstalling both updates. First, we uninstall KB 2667402:

 

Then we reboot the server.

 

Next, we uninstall KB 2621440:

 

Then we reboot the server again.

 

Now, we need to reinstall the updates, but we need to do so in this particular order.

 

First, install KB 2667402:

 

Then reboot the server.

 

Next, we install KB 2621440:

 

Then reboot the server again.