My website is not accessible anymore

coralsaw · 03-01-2006, 09:54 AM

Hello,

Please help me troubleshoot (Debian Sid), I have no idea what has changed since the last website test I did from the outside was a long time ago.

- I can ssh to my box from the outside, both by IP and domain.
- I can see the website from the box, but not from the outside.
- There is no firewall, and the router always routes everything to the one PC on the lan where the website (Apache2) is

Here's one strange thing though, the IP of the eth0 interface on ifconfig is not 192.168.7.13 as I would expect. It's different, and I have no idea why. Check it out:

Code:

eth0      Link encap:Ethernet  HWaddr 00:0C:6E:1F:6D:C5
          inet addr:169.254.187.182  Bcast:0.0.0.0  Mask:255.255.0.0
          inet6 addr: fe80::20c:6eff:fe1f:6dc5/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:36726 errors:0 dropped:0 overruns:0 frame:0
          TX packets:30349 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:44342571 (42.2 MiB)  TX bytes:3502445 (3.3 MiB)
          Interrupt:225

Here's also my hosts file (masked):

Code:

127.0.0.1       localhost.localdomain   localhost
192.168.7.13    studio.xxx.com    studio
192.168.7.1     router.xxx.com    router

# The following lines are desirable for IPv6 capable hosts
::1     ip6-localhost ip6-loopback
fe00::0 ip6-localnet
ff00::0 ip6-mcastprefix
ff02::1 ip6-allnodes
ff02::2 ip6-allrouters
ff02::3 ip6-allhosts

Also the route output:

Code:

Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
192.168.7.0     *               255.255.255.0   U     0      0        0 eth0
169.254.0.0     *               255.255.0.0     U     0      0        0 eth0
default         router.xxx.com  0.0.0.0         UG    0      0        0 eth0

Nothing shows in the apache log, router log or messages log when I try to hit the webserver from the outside.

Can you please help me?

TiA

/coralsaw

dracae · 03-01-2006, 10:06 AM

169.254.187.182 is given by some zeroconfig app. It happens when the link is up but the dhcp server cannot be contacted. Apt-cache suggests zcip

johnMG · 03-01-2006, 11:27 AM

> 169.254.187.182 is given by some zeroconfig app.

dracae, my /etc/network/interfaces file just says to use dhcp for eth0. If I restart networking, and there's no dhcp server running, my machine just tries and tries a number of times to get an address, then gives up.

Are you saying that some other program on coralsaw's machine is setting his system's IP address after his dhcp client gives up?

coralsaw, I'm guessing your /etc/network/interfaces just says

Code:

auto eth0
iface eth0 inet dhcp

Do you have any zeroconf packages installed? Try 'dpkg -l | grep zero'

coralsaw · 03-01-2006, 12:22 PM

Yes I do. Output:
ii zeroconf 0.6.1-1 IPv4 link-local address allocator

Also, /etc/network/interfaces says (for eth0)

# The primary network interface
auto eth0
iface eth0 inet dhcp

Should I not be able to ssh to the box if there was an IP problem? It seems like a hostname-related issue to me, although I would very much like to know why eth0 has that IP.

In anycase I purged zeroconf, I didn't need it anyway.

Any ideas?

TiA

/coralsaw

dracae · 03-01-2006, 12:33 PM

Quote:

Originally Posted by johnMG

>
Are you saying that some other program on coralsaw's machine is setting his system's IP address after his dhcp client gives up?
zero'

Yes, his dhcp server is not answering the request so zeroconf assumes that he is connecting to a hub/switch/crossover that does not have a dhcp server running. It will give out an ip in the 169.254.0.0/16 range. This way it can communicate with other machines supporting zeroconf, i.e. crossover cable connected to winxp workstation.

Either dhcp is not working/blocked or zeroconf doesn't know any better. Easiest solution would be to remove zeroconf, unless you are on a laptop AND have a valid use for it. If you then cannot get an address in your 192.168.7.0/24 range, then proceed from there. Unless you have static routes it looks like dhcp is working because you have routes on that network setup but the zeroconf routes have been added after that. What ever method zeroconf uses to check whether it is needed or not appears to be borked.

coralsaw · 03-02-2006, 02:15 AM

I purged zeroconf.

Altough route and ifconfig return expected results now and I can see the site from the box itself (either via domain name or via localhost:80), I still cannot see it from the outside.

The router carries all traffic to the box, which is why I can ssh to it, but for some weird reason, I cannot see the site, either via IP or domain name.

Any ideas? This is weird...

H2O-linux · 03-05-2006, 02:24 AM

What port is apache set to listen on? Maybe your ISP decided to blocked 80? Change it to 8080 in the /etc/httpd.conf file and " apachectl restart " . Make sure to add (port 8080) http://yourwebserver.com:8080 to see if it is seen outside the your home net.

johnMG · 03-07-2006, 11:58 AM

On my system, that file to edit is: /etc/apache2/ports.conf

Also, Apache 2 on Debian uses /etc/apache2/apache2.conf rather than httpd.conf (which is just present for backward compatibility reasons).