Jameson wrote:
> my xinetd for tftp looks like:
> service tftp
> { ...
>         server                  = /usr/sbin/in.tftpd
>         server_args             = -s /var/tftpboot -v
> ... }
>
> The howto I origianlly folowed had me using rc.local...

Hmmm...so one is presumably running constantly as a daemon and one is 
being launched by xinetd. I guess I'd recommend setting up both 
consistently, whichever approach you chose.


>> Are you sure you are using the correct high numbered port?
> 
> /usr/sbin/in.tftpd -l -a :16869 -s /var/tftpboot -v

The port number looks right. My BSD ported TFTP daemon doesn't document 
any switches other than -s and -n, so I can't say whether the syntax is 
correct. -l for listen and -a for address seems logical.

In theory, it should be safe to assume it is correct, as you said you 
tested connecting to the alternate port and it worked. But when you 
performed those tests, how confident are you that the tftp client was 
paying attention to the port you were specifying? Did you try connecting 
to a port you know has no TFTP daemon listening on it to see if it failed?

Note from your earlier logs:

> Dec 24 23:24:15 server xinetd[2126]: START: tftp pid=2954 from= 192.168.1.55
> Dec 24 23:25:51 server in.tftpd[2985]: RRQ from 192.168.1.55 filename 
> dongle.bin.ver
> Dec 24 23:27:07 server xinetd[3005]: START: tftp pid=3009 from=192.168.1.55
> Dec 24 23:27:07 server in.tftpd[3010]: RRQ from 192.168.1.55 filename 
> dongle.bin.ver
> Dec 24 23:42:07 server xinetd[3005]: EXIT: tftp status=0 pid=3009 
> duration=900(sec)

Is that suppose to be showing the logs from connections to both ports? 
It appears to be logging two file transfers that both went through a 
TFTP service launched by xinetd. That suggests you never successfully 
connected to the stand-alone daemon running on the high numbered port.

Another test to try is to point each TFTP server at a different 
directory, and put some unique files in each to test against. (You could 
even do this in "production" as the dongle and ver file need to be 
accessible from the high-numbered port, while the config file (startup 
script) and theme are pulled from the standard port.)

Also, do some sanity checks on the transferred file, like checking the 
size and checksum.


>> Which tftp server are you using?
>  
> Just the standard tftp-server in Fedora 8.

It appears it isn't atftpd nor the commonly available BSD port, but 
probably an enhanced derivative of the BSD port.


> The real problem is just that I never see it even trying to 
> hit my tftp server.

If none of the above pans out, I think you're going to have to break out 
tcpdump or wireshark.

  -Tom


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Mvpmc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mvpmc-users
mvpmc wiki: http://mvpmc.wikispaces.com/

Reply via email to