Ray,
Thank you for continuing this thread, and for the links you posted. However, I had thoroughly googled this problem before posting and there does not seem to be an *obvious* solution. I agree that there seems to be a "connection" problem - but I think it is in SME's firewall NAT code - either ingoing or outgoing, rather than, say, a problem in the code for the pptpd daemon. I rather suspect it is in the NATting of the outgoing packets from desktop client through my personal SME 6.01 server.
All the SME configurations are "stock" in terms of the firewall configurations and pptpd connection parameters. There are no "extra" filtering rules in the firewall configurations.
Here are some additional links I abstracted from the 'net:
http://lists.slug.org.au/archives/slug/2004/01/msg00585.html- the thread suggests that the firewall or ISP is blocking GRE packets
- author confirms that he can connect to another pptpd server with identical pptpd configuration options
http://ccfaq.valar.co.uk/modules.php?name=News&file=article- author(s) are using a different distribution - clarkconnect, but clarkconnect uses the same PoPToP pptpd setup, AFAIK
- author suggests using updated rpms from
www.poptop.orghttp://lists.contribs.org/mailman/public/devinfo/msg06683.html- author posted to devinfo with the same problem on SME, was redirected to bugs list
http://lists.netfilter.org/pipermail/netfilter/2003-June/045107.htmlhttp://lists.netfilter.org/pipermail/netfilter-devel/2003-September/012306.html- threads suggest that the problem is related to firewall NAT
http://sourceforge.net/mailarchive/message.php?msg_id=8990474- suggests that the incoming packets are being blocked, use tcpdump on server to analyse incoming packets
http://pptpclient.sourceforge.net/howto-diagnosis.phtml#write_epermSymptom: write to the GRE socket fails with EPERM.
Diagnosis: iptables rules (such as for a firewall configuration) do not allow the interface to emit GRE packets.
Solution: locate the rule that prevents the write, or add a rule to cover it, and retry the tunnel.
To test the possibility that it is the "outgoing" packets from my desktop machine going through my SME firewall that are related to the problem, I disconnected from my DSL network and used a dial-up account (from the same ISP) to connect to various remote SME servers. Result: of seven tests, two of which reliably connected through my SME server, six connected properly via the dial-up account. That indicates to me that the outbound NAT may be the source of the problem.
I think that the only way to truly identify the cause of this problem is to analyse the traffic between the SME servers involved in the connection. My understanding is than one can use "tcpdump" to dump the connection traffic, but while I have seen some instructions on how to use tcpdump to capture the traffic on the target server, I am unsure of how to dump the traffic on the originating server. Is there a tcpdump guru out there who is willing to help?
Peter