RPC unavailable / Outlook timeout

Sep 14 05 8:22 pm

Hi,

We are using Wingate VPN to host several VPN's. On the remote site of the VPN one is using Outlook connected to a Windows 2003 Exchange Server in local network (the host site of the VPN), a telnet application connected to a host in the local network and a few shared folders containing files, also on a Windows 2003 server in the local network. All is working fine, but after two or three weeks Outlook is timing out and I can't 'dir' a larger share (error 1726, rpc service unavailable). The telnet application is still working fine and I'm still able to ping everything across the VPN's.

Solving this problem is easy; by restarting the Windows 2003 server everything is ok. This morning I tried changing MTU's on a joining site, I went down all the way to 330, but this didn't seem to solve the problem. But now I even can't take over the remote Wingate-pc with VNC ...!

What's going on?

Thanks, Bas.

Sep 14 05 8:31 pm

When you were changing the MTU, were you using the MTU option in GateKeeper? Can you try setting it with Dr. TCP? http://www.dslreports.com/drtcp
Also, here is a post with instructions on doing the Ping test: http://forums.qbik.com/viewtopic.php?p=18500#18500

Sep 14 05 9:02 pm

Hi James,

I changed the MTU in Gatekeeper, the setting seems to be active only after a reboot. I cannot connect to the pc anymore because VNC fails to load the initial screen.

Can I try setting the MTU on the host-pc?

I tried the ping-test, initially 1500 was ok, then I changed the MTU on the joining pc to 800. Result: pinging with 800 fails and I had to go down to 330!

Can you explain why things are working fine for two-three weeks? And why rebooting the Windows 2003 server (not the Wingate-pc) solves the problem?

Bas.

Sep 14 05 9:13 pm

Can I try setting the MTU on the host-pc?

Yes

I tried the ping-test, initially 1500 was ok, then I changed the MTU on the joining pc to 800. Result: pinging with 800 fails and I had to go down to 330!

Gee, that is low

Can you explain why things are working fine for two-three weeks? And why rebooting the Windows 2003 server (not the Wingate-pc) solves the problem?

No, but I have sent an email to one of the developers regarding this question; one of us will advise.

1. Also, were you setting the MTU with Dr. TCP?
2. Do you have a windows service pack on your Win2003 box?

Sep 14 05 9:21 pm

Also, can you give the basic layout of this problem:

e.g.
VPN Joiner at remote site (Win2000)
|
VPN Host at main office (Win 2000)
|
Exchange server 2005 on main office network (Win 2003 no SP)

Sep 14 05 11:21 pm

Hi James,

The layout is as follows;

VPN Host and Joiner:
W2K SP4
2 NIC's (one to local network, one connected to ADSL)

Workstations on remote LAN:
W2K SP4 and WXP SP?

Server on local network:
W2K3 SBS Server with SP1
Exchange v6.5 SP1

In the mean time I have reset the gateway I couldn't reach to MTU 800 and it works, now I can take over the pc with VNC. Since I restarted the server this morning Outlook is working fine.

I've not used DrTCP, I changed the settings with Gatekeeper.

Bas.

Sep 15 05 6:39 pm

The developer has reviewed this post, and has asked you to set the MTU via Dr. TCP. We have discovered there are some issues with changing the MTU via the GateKeeper interface; that will be fixed in the next release. If you are still having problems, please use Dr. TCP to change the MTU.

http://www.dslreports.com/drtcp

Regarding the rebooting of the server fixes the problem. We cannot give a definitive answer to this but can speculate:
1. Rebooting the server sets it back to a default MTU.
2. This could be caused by the ISP telling your connection to change the MTU
3. Could be an advanced feature of your network card that is interfering; e.g Jumbo Frames
4. Instead of rebooting the server, you may just need to disable/enable the network card.

Also, what version of WinGate are you using (or WinGateVPN Only)?