VPN Drops every 9 minutes

I have a client that lives in a rural area and connects to a wireless ISP, when she connects to the office via pptp vpn she gets disconnected every 9 minutes, give or take 30 seconds.

ISP uses Tranzeo radios. In her house she uses a wireless laptop connected to a wrt54gl running dd-wrt 23 sp2.

Her office is running server 2003 for remote access behind a sonicwall tz170.

To see if the problem was her laptop I tried using the dd-wrt pptp client to connect and left it run overnight. Server logs show drops approx every 9 minutes. I then tried using the dd-wrt client to connect to a different server at my office. Same thing the vpn drops every 9 minutes.

Something is kicking her off, the only thing I can think of is a weather radar station located a few miles away. Could this be the culprit? Anyone have any ideas as to what else could be causing this?

Ian

Reply to
Ian
Loading thread data ...

"Ian" hath wroth:

Is the PPTP server terminated in the Sonicwall or in the Windoze 2003 server?

Not enough info to make a conclusive guess. It could be literally any source of RF. See list of probable culprits at:

Offhand, I don't see anything on the list with a 9 minute cycle time.

I use the PPTP client and server in DD-WRT v23 SP2 and SP3 without any dropouts, so that's not it. There is the "scheduled reboot" feature, which may have accidentally been setup. See:

That's probably not the problem, but might be worth checking.

If the VPN is terminated in the Windoze 2003 server, look in the "Event Viewer" for clues as to which end initiated the disconnects and what was happening when it disconnected.

Also monitor connectivity during the critical 9 minute period. A simple ping running every 5 seconds should be sufficient. If the pings disappear at the same time as the VPN goes down, you've lost the RF link, probably due to RF intererence. If the pings continue normally when the VPN goes down, then it's something in either the PPTP client, server, or Sonicwall router.

Don't ignore the in between boxes. I had an allegedly working router drive me nuts by silently rebooting every 2 hours. No evidence of a reboot in syslog. I eventually noticed that the uptime was unusually low, which pointed to the reboots.

If it points to an RF problem, you'll probably need to buy or borrow some form of RF spectrum analyzer. Check with the local hams or try Wi-Spy at:

Ouch. $400. I paid $100 when it first was introduced.

Well, this one is $100 for the "lite" version, but I don't know anything about it.

Reply to
Jeff Liebermann

PPTP terminates in the server

I checked, the keep alive is disabled

Event viewer most of the time says "the reason for disconnecting was administrative settings or explicit request" but some times says "user request". Google doesn't help much with the first one.

Good idea, the pings continue when connectivity drops.

Since the same thing happens on two diffferent Windows servers in two different locations only one behind a sonicwall I'll concentrate on the client end, bypass router, change router etc. She is the only user this happens to on both servers.

One other thing I forgot to mention I tried forwarding terminal services ports and had her remote desktop to a system directly without VPN, she said it was no better.

Thanks alot for the help Jeff, looks like I'm straying away from wireless but will post if I resolve the issue.

Ian

Reply to
Ian

"Ian" hath wroth:

Also, check the Event Viewer on the client computah. It might have additional clues. The "user request" disconnect means that it was initiated from the client end. However, you get the same message if you just pull the plug on the ethernet cable, so that doesn't help. I couldn't find anything sane for the "administrative settings" message. It might be something in the Group Policy Management application, if it's being used. Hard to tell from here.

Good. That eliminates a large number of probable culprits. It's not connectivity, wireless, the Sonicwall, or a loose cable. It's down to either the PPTP server or the client.

If the servers are replicating each others security database (i.e. Backup Domain Controller), then they can be expected to act identically. I found a problem with a Windoze 2003 server today that might be relevent. The server would fill up the disk write cache with a huge amount of data to be written to disk. When it hit some threshold, it would flush the cache all at once, during which time, everything would come to a grinding halt. I noticed that I was getting disconnect/reconnect messages from the remote VPN connections. I haven't really solved the problem yet, so I don't have a suitable fix.

Well, that eliminates the two servers, leaving only the client computer. It still might be something exotic, but my guess is that the client is getting busied out by something, and causing disconnects due to timeouts.

OK. Good luck.

Reply to
Jeff Liebermann

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.