PDA

View Full Version : Daily Update over Network FAILS



larray
09-09-2003, 10:40 AM
Does anyone have a process flow for how the "dial out" over the network works? (IP's, PORTS, logs, etc)

My current setup -
Tivo SA S1 with Turbonet

All networking functions properly to the box - I can see the Tivo on the wire, telnet, FTP, mfs_ftp - all work fine.

I changed the firewall/router that I'm using to gain some extra features (logging, filtering, etc), and now the Tivo can't call home. Test call and full call fail.

If I put the SMC router or a DLink router (residential gateway) back in as the router between my DSL connection and my home network, BAM the Tivo has no problems. Put the laptop router back in place, no connection.

My firewall rules are setup to allow all internal traffic access to the outside wire, but I'm not even getting any packet traffic from the Tivo's address.

I saw this post here:
http://www.dealdatabase.com/forum/showthread.php?threadid=23302&highlight=tracert

that alluded to an http file in /var/log which I cannot find on a S1 SA (it appears to be Dtivo specific since I can find it on a DTivo). A log of the call home process would be very helpful, otherwise it looks like I'm back to sniffing, going back to PPP over serial (sice I think it did generate a log), or putting the old router back in place.

TIA

larray
09-10-2003, 09:42 AM
Ok - here's a bit more info that I've found for anyone who may encounter problems with their daily call over the network / ethernet.

I FINALLY found a log that somewhat pertains to the daily call - it appears to cover the packaging and dialout proceedure and I get a tiny snippit of the problem. The log is /var/log/tclient and Otclient.

The interesting data is:


Sep 10 01:25:09 (none) comm[126]: Backdoor code 4, enable calls via broadband
Sep 10 01:25:09 (none) comm[126]: Using Ethernet. Not starting modem/pppd.
Sep 10 01:25:09 (none) comm[126]: CallStatusReporter: Phase: Start_Auth, Status In Progress
Sep 10 01:25:09 (none) comm[126]: CallStatusReporter: sending message "ST|33"
Sep 10 01:25:09 (none) comm[126]: CommUtil: connection to host 204.176.49.2, port 80, err 0x0
Sep 10 01:25:09 (none) comm[126]: Uploading HTTP Header for modLog of /var/log/svclog: POST /tivo-service/mlog.cgi HTTP/1.0^M Content-Length: 2769^M ^M
Sep 10 01:25:09 (none) comm[126]: Bad HTTP response: HTTP/1.0 302 Redirect
Sep 10 01:25:09 (none) comm[126]: XferRqst Failed to write all bytes: Bad HTTP response
Sep 10 01:25:09 (none) comm[126]: drainGetPostQ: doXfer failed err=65535 (0xffff)
Sep 10 01:25:09 (none) comm[126]: NetAgent: drainPostQ failed, err = 0xffff
Sep 10 01:25:09 (none) comm[126]: CallService: NetAgent Process err 65535
Sep 10 01:25:09 (none) comm[126]: CallService: NetAgent failed, status 0xffff
Sep 10 01:25:09 (none) comm[126]: CallService: EndCall start
Sep 10 01:25:11 (none) comm[126]: CallStatusReporter: Turning off-hook LED off


There isn't a whole lot of data on a google search for phrases about "Bad HTTP Response" or "302 redirect". Any one a little more web savvy or networking savvy that might be able to help it would be appreciated.

FWIW - the setup I have is through a dedicated routing laptop with Kerio firewall on it. Internal lan traffic is allowed unrestricted access to the DSL connection.

The tivo has no problem pinging 204.176.49.2. Best I can tell the tivo initiates a connection to 204.176.49.2:80 from it's port 1092 (that may change, but I just started capturing packets on the firewall from the tivo last night) I see some traffic flow, then on the firewall it abruptly ends.

Anyone want a virtual beer for the solution to the perplexing problem?