a few network problems

Hi,

I’m having a few network problems that I would appreciate some pointers
on solutions.

I have a server running debian linux, which has been working for a few
months now. But I have recently encountered a few problems. The server
is connected to the internet via an ADSL modem and gets its IP via DHCP
from my ISP. Connected to it on the internal network are two PC running
windows XP and a due boot laptop running Linux / windows XP.

The problems I am having are:

1. Losing IP number. When I boot up I get an IP number from my ISP with
a time out of 400 – 600 seconds. My connection to the internet lasts for
about 5 minutes then I lose the connection. When I run ifconfig I see I
have lost my IP address provided by my ISP and have an address have an
address starting with 169. This problem started on Saturday evening but
cleared up itself on Sunday morning. Things were working fine again
until Wednesday evening when the problem reoccurred. This made me think
it was a problem with my ISP but they didn’t seam to have a clue what
the problem could be. So I check if the windows machine running XP can
connect and it can. No problems in windows. What I would like to know
is; is this a problem with my server or my ISP Why does it work under
windows but not linux? If the problem is with my sever, how do I find
out what the problem is and fix it?

2. This is related to the first problem. When I lose connection to my
ISP I also lose connection to all computers on the internal network. I
get “network unreachable” errors even with the server. I thought this
might be a DNS problem as I’m using my ISP’s DNS. However, after
setting up DNS on my server and using that the problem remains. Any
ideas what could cause it and how to fix it?

3. This problem is an old problem that I have had since first setting up
the server. I have set up samba to allow directory sharing across the
network. I can access shared directories on the windows machine from
linux with no problems. I can access shared directories on a linux
machine from a windows machine. However, I can not access shared
directories on windows machines from a windows machine. Also, sometimes
I can see the other machines on the network from a windows machine and
sometimes I can’t. On one occasion I could see all the other computers
on the network from a windows machine and as I was looking all the other
computer vanished! As things work from the linux end, am I correct in
assuming the problem is with windows? If so, how do I go about fixing it?

Thanks for your time.

AAW wrote:
> Hi,
>
> I’m having a few network problems that I would appreciate some pointers
> on solutions.
>
> I have a server running debian linux, which has been working for a few
> months now. But I have recently encountered a few problems. The server
> is connected to the internet via an ADSL modem and gets its IP via DHCP
> from my ISP. Connected to it on the internal network are two PC running
> windows XP and a due boot laptop running Linux / windows XP.
>
> The problems I am having are:
>
> 1. Losing IP number. When I boot up I get an IP number from my ISP
> with a time out of 400 – 600 seconds. My connection to the internet
> lasts for about 5 minutes then I lose the connection. When I run
> ifconfig I see I have lost my IP address provided by my ISP and have an
> address have an address starting with 169. This problem started on
> Saturday evening but cleared up itself on Sunday morning. Things were
> working fine again until Wednesday evening when the problem reoccurred.
> This made me think it was a problem with my ISP but they didn’t seam to
> have a clue what the problem could be. So I check if the windows machine
> running XP can connect and it can. No problems in windows. What I would
> like to know is; is this a problem with my server or my ISP Why does it
> work under windows but not linux? If the problem is with my sever, how
> do I find out what the problem is and fix it?
>
My first resort would be to do a reinstall of the dhcp client software.
Also, There will be log messages somewhere in /var/log/ (daemon.log on
my firewall) giving details of the link state and any negotiation messages.
> 2. This is related to the first problem. When I lose connection to my
> ISP I also lose connection to all computers on the internal network. I
> get “network unreachable” errors even with the server. I thought this
> might be a DNS problem as I’m using my ISP’s DNS. However, after
> setting up DNS on my server and using that the problem remains. Any
> ideas what could cause it and how to fix it?
>
This part should be independent of 1 unless the dhcp client is upsetting
things. First call would be the output of ifconfig and route -n under
both conditions (with and without the IP), I'd also check
/etc/network/interfaces
/etc/[network|ppp]/ip-[down|up].d/
and any scripts associated with your modem driver.

--
Keith Blow

Keith,

Thanks. Sounds like a good start. I will have a go when I get back home.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


>
> 3. This problem is an old problem that I have had since first setting
> up the server. I have set up samba to allow directory sharing across the
> network. I can access shared directories on the windows machine from
> linux with no problems. I can access shared directories on a linux
> machine from a windows machine. However, I can not access shared
> directories on windows machines from a windows machine. Also, sometimes
> I can see the other machines on the network from a windows machine and
> sometimes I can’t. On one occasion I could see all the other computers
> on the network from a windows machine and as I was looking all the other
> computer vanished! As things work from the linux end, am I correct in
> assuming the problem is with windows? If so, how do I go about fixing it?
>
Hello Andrew:
first of all you shlould think that the user from one windows machine
must exist on the other machine, both users must have the same
passwords, otherwise you will experience a strange beahvior on this.

as a next step you should test the following: ping one machine to the
tother. is it working?
then i would test the wins/dns resolution:

take the computername from the other machine and make:
ping workstation1 (f.i.) this should work on any windows network. if
this is not working you can call the computer by name or ip from the
point run in the windows panel with:
\\workstation1
\\ip-address
in both cases you should see an exact errormessage. if the
networkneighboorhodd is not working or not browseable so you can access
the shared folders in most cases.
for further investigations you should have a look in the loggings from
the samba server, sometimes there are interesting loggings to find.
/var/log/samba/smbd.conf

Greetings Randolf Balasus







-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHPCfkPtLQAXC/xvYRAuwXAJ9RlhUHqGnNYmk5RB+vrtmvTavKoQCfY8Ts
na6J2KuHZLHOez0lXFYlGFY=
=Lhm2
-----END PGP SIGNATURE-----

Randolf,


<<first of all you shlould think that the user from one windows machine
> must exist on the other machine, both users must have the same
> passwords, otherwise you will experience a strange beahvior on this.>>

Yes, the user exists and I can ping the other machines.

<<then i would test the wins/dns resolution:>>

That bit I haven't check so will do.

Thanks.


Randolf Balasus wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
>
>> 3. This problem is an old problem that I have had since first setting
>> up the server. I have set up samba to allow directory sharing across the
>> network. I can access shared directories on the windows machine from
>> linux with no problems. I can access shared directories on a linux
>> machine from a windows machine. However, I can not access shared
>> directories on windows machines from a windows machine. Also, sometimes
>> I can see the other machines on the network from a windows machine and
>> sometimes I can’t. On one occasion I could see all the other computers
>> on the network from a windows machine and as I was looking all the other
>> computer vanished! As things work from the linux end, am I correct in
>> assuming the problem is with windows? If so, how do I go about fixing it?
>>
> Hello Andrew:
> first of all you shlould think that the user from one windows machine
> must exist on the other machine, both users must have the same
> passwords, otherwise you will experience a strange beahvior on this.
>
> as a next step you should test the following: ping one machine to the
> tother. is it working?
> then i would test the wins/dns resolution:
>
> take the computername from the other machine and make:
> ping workstation1 (f.i.) this should work on any windows network. if
> this is not working you can call the computer by name or ip from the
> point run in the windows panel with:
> \\workstation1
> \\ip-address
> in both cases you should see an exact errormessage. if the
> networkneighboorhodd is not working or not browseable so you can access
> the shared folders in most cases.
> for further investigations you should have a look in the loggings from
> the samba server, sometimes there are interesting loggings to find.
> /var/log/samba/smbd.conf
>
> Greetings Randolf Balasus
>
>
>
>
>
>
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.6 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFHPCfkPtLQAXC/xvYRAuwXAJ9RlhUHqGnNYmk5RB+vrtmvTavKoQCfY8Ts
> na6J2KuHZLHOez0lXFYlGFY=
> =Lhm2
> -----END PGP SIGNATURE-----

I did an uninstall and a reinstall but no change. So I have had a look
at the logs. I can see that it gets its IP address from my ISP :

Nov 15 15:46:04 skamis dhclient: bound to 213.66.199.118 -- renewal in
530 seconds.

and then I can see that it loses its address :

Nov 15 15:46:44 skamis dhclient: DHCPRELEASE on eth0 to 213.66.199.1 port 67

....

Iautoip: Sending probe #0 for IP address 169.254.81.0.


but I don't really follow why. One thing I note in the log is the bit
about "Avahi detected that your currently configured local DNS server
serves a domain .local" leading to "Avahi disabled itself". Could this
be the source of the problem?

Part of the log is bellow. Does it help to work out why I lose my IP
address?

Thanks for your time.

log -----

Nov 15 15:45:50 skamis dhclient: Internet Systems Consortium DHCP Client
V3.0.4
Nov 15 15:45:50 skamis dhclient: Copyright 2004-2006 Internet Systems
Consortium.
Nov 15 15:45:50 skamis dhclient: All rights reserved.
Nov 15 15:45:50 skamis dhclient: For info, please visit
http://www.isc.org/sw/dhcp/
Nov 15 15:45:50 skamis dhclient:
Nov 15 15:45:50 skamis dhclient: Listening on LPF/eth0/00:04:75:dc:7e:9b
Nov 15 15:45:50 skamis dhclient: Sending on LPF/eth0/00:04:75:dc:7e:9b
Nov 15 15:45:50 skamis dhclient: Sending on Socket/fallback
Nov 15 15:45:50 skamis dhclient: DHCPRELEASE on eth0 to 213.66.199.1 port 67
Nov 15 15:45:50 skamis dhclient: send_packet: Network is unreachable
Nov 15 15:45:50 skamis dhclient: send_packet: please consult README file
regarding broadcast address.
Nov 15 15:45:50 skamis avahi-daemon[4047]: Withdrawing address record
for 169.254.81.0 on eth0.
Nov 15 15:45:50 skamis dhclient: receive_packet failed on eth0: Network
is down
Nov 15 15:45:50 skamis avahi-daemon[4047]: Leaving mDNS multicast group
on interface eth0.IPv4 with address 169.254.81.0.
Nov 15 15:45:50 skamis avahi-daemon[4047]: iface.c:
interface_mdns_mcast_join() called but no local address available.
Nov 15 15:45:50 skamis avahi-daemon[4047]: Interface eth0.IPv4 no longer
relevant for mDNS.
Nov 15 15:45:50 skamis NetworkManager: <information>^ISWITCH:
terminating current connection 'eth0' because it's no longer valid.
Nov 15 15:45:50 skamis NetworkManager: <information>^IDeactivating
device eth0.
Nov 15 15:45:50 skamis NetworkManager: nm_device_is_802_3_ethernet:
assertion `dev != NULL' failed
Nov 15 15:45:50 skamis NetworkManager: nm_device_is_802_11_wireless:
assertion `dev != NULL' failed
Nov 15 15:45:58 skamis dhclient: Internet Systems Consortium DHCP Client
V3.0.4
Nov 15 15:45:58 skamis dhclient: Copyright 2004-2006 Internet Systems
Consortium.
Nov 15 15:45:58 skamis dhclient: All rights reserved.
Nov 15 15:45:58 skamis dhclient: For info, please visit
http://www.isc.org/sw/dhcp/
Nov 15 15:45:58 skamis dhclient:
Nov 15 15:45:58 skamis kernel: eth0: setting full-duplex.
Nov 15 15:45:58 skamis dhcdbd: message_handler: message handler not
found under /com/redhat/dhcp/eth0 for sub-path eth0.dbus.get.reason
Nov 15 15:45:58 skamis NetworkManager: <information>^IWill activate
wired connection 'eth0' because it now has a link.
Nov 15 15:45:58 skamis NetworkManager: <information>^ISWITCH: no current
connection, found better connection 'eth0'.
Nov 15 15:45:58 skamis dhcdbd: message_handler: message handler not
found under /com/redhat/dhcp/eth0 for sub-path eth0.dbus.get.reason
Nov 15 15:45:58 skamis NetworkManager: <information>^IWill activate
connection 'eth0'.
Nov 15 15:45:58 skamis NetworkManager: <information>^IDevice eth0
activation scheduled...
Nov 15 15:45:58 skamis NetworkManager: <information>^IActivation (eth0)
started...
Nov 15 15:45:58 skamis NetworkManager: <information>^IActivation (eth0)
Stage 1 of 5 (Device Prepare) scheduled...
Nov 15 15:45:58 skamis NetworkManager: <information>^IActivation (eth0)
Stage 1 of 5 (Device Prepare) started...
Nov 15 15:45:58 skamis NetworkManager: <information>^IActivation (eth0)
Stage 2 of 5 (Device Configure) scheduled...
Nov 15 15:45:58 skamis NetworkManager: <information>^IActivation (eth0)
Stage 1 of 5 (Device Prepare) complete.
Nov 15 15:45:58 skamis NetworkManager: <information>^IActivation (eth0)
Stage 2 of 5 (Device Configure) starting...
Nov 15 15:45:58 skamis NetworkManager: <information>^IActivation (eth0)
Stage 2 of 5 (Device Configure) successful.
Nov 15 15:45:58 skamis NetworkManager: <information>^IActivation (eth0)
Stage 3 of 5 (IP Configure Start) scheduled.
Nov 15 15:45:58 skamis NetworkManager: <information>^IActivation (eth0)
Stage 2 of 5 (Device Configure) complete.
Nov 15 15:45:58 skamis NetworkManager: <information>^IActivation (eth0)
Stage 3 of 5 (IP Configure Start) started...
Nov 15 15:45:59 skamis NetworkManager: <information>^IActivation (eth0)
Beginning DHCP transaction.
Nov 15 15:45:59 skamis NetworkManager: <information>^IActivation (eth0)
Stage 3 of 5 (IP Configure Start) complete.
Nov 15 15:45:59 skamis NetworkManager: <information>^IDHCP daemon state
is now 12 (successfully started) for interface eth0
Nov 15 15:45:59 skamis dhclient: Listening on LPF/eth0/00:04:75:dc:7e:9b
Nov 15 15:45:59 skamis dhclient: Sending on LPF/eth0/00:04:75:dc:7e:9b
Nov 15 15:45:59 skamis dhclient: Sending on Socket/fallback
Nov 15 15:46:01 skamis dhclient: DHCPDISCOVER on eth0 to 255.255.255.255
port 67 interval 4
Nov 15 15:46:01 skamis dhclient: DHCPOFFER from 213.66.199.1
Nov 15 15:46:01 skamis dhclient: DHCPREQUEST on eth0 to 255.255.255.255
port 67
Nov 15 15:46:01 skamis dhclient: DHCPACK from 213.66.199.1
Nov 15 15:46:01 skamis avahi-daemon[4047]: New relevant interface
eth0.IPv4 for mDNS.
Nov 15 15:46:01 skamis avahi-daemon[4047]: Joining mDNS multicast group
on interface eth0.IPv4 with address 213.66.199.118.
Nov 15 15:46:01 skamis avahi-daemon[4047]: Registering new address
record for 213.66.199.118 on eth0.
Nov 15 15:46:01 skamis dhclient: bound to 213.66.199.118 -- renewal in
525 seconds.
Nov 15 15:46:01 skamis avahi-daemon[4047]: Got SIGTERM, quitting.
Nov 15 15:46:01 skamis avahi-daemon[4047]: Leaving mDNS multicast group
on interface eth3.IPv4 with address 192.168.3.103.
Nov 15 15:46:01 skamis avahi-daemon[4047]: Leaving mDNS multicast group
on interface eth2.IPv4 with address 192.168.2.102.
Nov 15 15:46:01 skamis avahi-daemon[4047]: Leaving mDNS multicast group
on interface eth1.IPv4 with address 192.168.1.101.
Nov 15 15:46:01 skamis avahi-daemon[4047]: Leaving mDNS multicast group
on interface eth0.IPv4 with address 213.66.199.118.
Nov 15 15:46:01 skamis avahi: Avahi detected that your currently
configured local DNS server serves
Nov 15 15:46:01 skamis avahi: a domain .local. This is inherently
incompatible with Avahi and thus
Nov 15 15:46:01 skamis avahi: Avahi disabled itself. If you want to use
Avahi in this network, please
Nov 15 15:46:01 skamis avahi: contact your administrator and convince
him to use a different DNS domain,
Nov 15 15:46:01 skamis avahi: since .local should be used exclusively
for Zeroconf technology.
Nov 15 15:46:01 skamis avahi: For more information, see
http://avahi.org/wiki/AvahiAndUnicastDotLocal
Nov 15 15:46:02 skamis avahi: Avahi detected that your currently
configured local DNS server serves
Nov 15 15:46:02 skamis avahi: a domain .local. This is inherently
incompatible with Avahi and thus
Nov 15 15:46:02 skamis avahi: Avahi disabled itself. If you want to use
Avahi in this network, please
Nov 15 15:46:02 skamis avahi: contact your administrator and convince
him to use a different DNS domain,
Nov 15 15:46:02 skamis avahi: since .local should be used exclusively
for Zeroconf technology.
Nov 15 15:46:02 skamis avahi: For more information, see
http://avahi.org/wiki/AvahiAndUnicastDotLocal
Nov 15 15:46:04 skamis dhclient: DHCPDISCOVER on eth0 to 255.255.255.255
port 67 interval 8
Nov 15 15:46:04 skamis dhclient: DHCPOFFER from 213.66.199.1
Nov 15 15:46:04 skamis dhclient: DHCPREQUEST on eth0 to 255.255.255.255
port 67
Nov 15 15:46:04 skamis dhclient: DHCPACK from 213.66.199.1
Nov 15 15:46:04 skamis dhclient: bound to 213.66.199.118 -- renewal in
530 seconds.
Nov 15 15:46:08 skamis kernel: eth0: no IPv6 routers present
Nov 15 15:46:13 skamis dhclient: DHCPREQUEST on eth0 to 213.66.199.1 port 67
Nov 15 15:46:13 skamis dhclient: DHCPACK from 213.66.199.1
Nov 15 15:46:13 skamis dhclient: bound to 213.66.199.118 -- renewal in
541 seconds.
Nov 15 15:46:44 skamis NetworkManager: <information>^IDevice 'eth0' DHCP
transaction took too long (>45s), stopping it.
Nov 15 15:46:44 skamis dhclient: There is already a pid file
/var/run/dhclient.eth0.pid with pid 5704
Nov 15 15:46:44 skamis dhclient: killed old client process, removed PID file
Nov 15 15:46:44 skamis dhclient: DHCPRELEASE on eth0 to 213.66.199.1 port 67
Nov 15 15:46:45 skamis NetworkManager: <information>^IActivation (eth0)
Stage 4 of 5 (IP Configure Timeout) scheduled...
Nov 15 15:46:45 skamis NetworkManager: <information>^IDHCP daemon state
is now 14 (normal exit) for interface eth0
Nov 15 15:46:45 skamis NetworkManager: <information>^IDHCP daemon state
is now 14 (normal exit) for interface eth0
Nov 15 15:46:45 skamis NetworkManager: <information>^IActivation (eth0)
Stage 4 of 5 (IP Configure Timeout) started...
Nov 15 15:46:45 skamis NetworkManager: <information>^INo DHCP reply
received. Automatically obtaining IP via Zeroconf.
Nov 15 15:46:45 skamis NetworkManager: <information>^Iautoip: Sending
probe #0 for IP address 169.254.81.0.
Nov 15 15:46:45 skamis NetworkManager: <information>^Iautoip: Waiting
for reply...
Nov 15 15:46:47 skamis NetworkManager: <information>^Iautoip: Sending
probe #1 for IP address 169.254.81.0.
Nov 15 15:46:47 skamis NetworkManager: <information>^Iautoip: Waiting
for reply...
Nov 15 15:46:49 skamis NetworkManager: <information>^Iautoip: Sending
probe #2 for IP address 169.254.81.0.
Nov 15 15:46:49 skamis NetworkManager: <information>^Iautoip: Waiting
for reply...
Nov 15 15:46:52 skamis NetworkManager: <information>^Iautoip: Sending
announce #0 for IP address 169.254.81.0.
Nov 15 15:46:52 skamis NetworkManager: <information>^Iautoip: Waiting
for reply...
Nov 15 15:46:55 skamis NetworkManager: <information>^Iautoip: Sending
announce #1 for IP address 169.254.81.0.
Nov 15 15:46:55 skamis NetworkManager: <information>^Iautoip: Waiting
for reply...
Nov 15 15:46:58 skamis NetworkManager: <information>^Iautoip: Sending
announce #2 for IP address 169.254.81.0.
Nov 15 15:46:58 skamis NetworkManager: <information>^Iautoip: Waiting
for reply...
Nov 15 15:47:01 skamis NetworkManager: <information>^IActivation (eth0)
Stage 5 of 5 (IP Configure Commit) scheduled...
Nov 15 15:47:01 skamis NetworkManager: <information>^IActivation (eth0)
Stage 4 of 5 (IP Configure Timeout) complete.
Nov 15 15:47:01 skamis NetworkManager: <information>^IActivation (eth0)
Stage 5 of 5 (IP Configure Commit) started...
Nov 15 15:47:02 skamis NetworkManager: <information>^IClearing nscd
hosts cache.
Nov 15 15:47:02 skamis NetworkManager: <WARNING>^I nm_spawn_process ():
nm_spawn_process('/usr/sbin/nscd -i hosts'): could not spawn process.
(Failed to execute child process "/usr/sbin/nscd" (No such file or
directory))
Nov 15 15:47:02 skamis NetworkManager: <information>^IActivation (eth0)
successful, device activated.
Nov 15 15:47:02 skamis NetworkManager: <information>^IActivation (eth0)
Finish handler scheduled.
Nov 15 15:47:02 skamis NetworkManager: <information>^IActivation (eth0)
Stage 5 of 5 (IP Configure Commit) complete.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Found user 'avahi' (UID 104)
and group 'avahi' (GID 105).
Nov 15 15:47:03 skamis avahi-daemon[5831]: Successfully dropped root
privileges.
Nov 15 15:47:03 skamis avahi-daemon[5831]: avahi-daemon 0.6.16 starting up.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Successfully called chroot().
Nov 15 15:47:03 skamis avahi-daemon[5831]: Successfully dropped
remaining capabilities.
Nov 15 15:47:03 skamis avahi-daemon[5831]: No service found in
/etc/avahi/services.
Nov 15 15:47:03 skamis avahi-daemon[5831]: New relevant interface
eth3.IPv4 for mDNS.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Joining mDNS multicast group
on interface eth3.IPv4 with address 192.168.3.103.
Nov 15 15:47:03 skamis avahi-daemon[5831]: New relevant interface
eth2.IPv4 for mDNS.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Joining mDNS multicast group
on interface eth2.IPv4 with address 192.168.2.102.
Nov 15 15:47:03 skamis avahi-daemon[5831]: New relevant interface
eth1.IPv4 for mDNS.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Joining mDNS multicast group
on interface eth1.IPv4 with address 192.168.1.101.
Nov 15 15:47:03 skamis avahi-daemon[5831]: New relevant interface
eth0.IPv4 for mDNS.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Joining mDNS multicast group
on interface eth0.IPv4 with address 169.254.81.0.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Network interface enumeration
completed.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Registering new address
record for 192.168.3.103 on eth3.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Registering new address
record for 192.168.2.102 on eth2.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Registering new address
record for 192.168.1.101 on eth1.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Registering new address
record for 169.254.81.0 on eth0.
Nov 15 15:47:03 skamis avahi-daemon[5831]: Registering HINFO record with
values 'I686'/'LINUX'.
Nov 15 15:47:04 skamis avahi-daemon[5831]: Server startup complete. Host
name is skamis.local. Local service cookie is 1032625664.
Nov 15 15:47:13 skamis kernel: eth0: no IPv6 routers present



----
AAW wrote:
> Hi,
>
> I’m having a few network problems that I would appreciate some
pointers on solutions.
>
> I have a server running debian linux, which has been working for a
few months now. But I have recently encountered a few problems. The
server is connected to the internet via an ADSL modem and gets its IP
via DHCP from my ISP. Connected to it on the internal network are two PC
running windows XP and a due boot laptop running Linux / windows XP.
>
> The problems I am having are:
>
> 1. Losing IP number. When I boot up I get an IP number from my ISP
with a time out of 400 – 600 seconds. My connection to the internet
lasts for about 5 minutes then I lose the connection. When I run
ifconfig I see I have lost my IP address provided by my ISP and have an
address have an address starting with 169. This problem started on
Saturday evening but cleared up itself on Sunday morning. Things were
working fine again until Wednesday evening when the problem reoccurred.
This made me think it was a problem with my ISP but they didn’t seam to
have a clue what the problem could be. So I check if the windows machine
running XP can connect and it can. No problems in windows. What I would
like to know is; is this a problem with my server or my ISP Why does it
work under windows but not linux? If the problem is with my sever, how
do I find out what the problem is and fix it?
>
My first resort would be to do a reinstall of the dhcp client software.
Also, There will be log messages somewhere in /var/log/ (daemon.log on
my firewall) giving details of the link state and any negotiation messages.
> 2. This is related to the first problem. When I lose connection to
my ISP I also lose connection to all computers on the internal network.
I get “network unreachable” errors even with the server. I thought this
might be a DNS problem as I’m using my ISP’s DNS. However, after
setting up DNS on my server and using that the problem remains. Any
ideas what could cause it and how to fix it?
>
This part should be independent of 1 unless the dhcp client is upsetting
things. First call would be the output of ifconfig and route -n under
both conditions (with and without the IP), I'd also check
/etc/network/interfaces
/etc/[network|ppp]/ip-[down|up].d/
and any scripts associated with your modem driver.

--
Keith Blow

AAW wrote:
> I did an uninstall and a reinstall but no change. So I have had a look
> at the logs. I can see that it gets its IP address from my ISP :
>
> Nov 15 15:46:04 skamis dhclient: bound to 213.66.199.118 -- renewal in
> 530 seconds.
>
> and then I can see that it loses its address :
>
> Nov 15 15:46:44 skamis dhclient: DHCPRELEASE on eth0 to 213.66.199.1
> port 67
>
> ...
>
> Iautoip: Sending probe #0 for IP address 169.254.81.0.
>
>
> but I don't really follow why. One thing I note in the log is the bit
> about "Avahi detected that your currently configured local DNS server
> serves a domain .local" leading to "Avahi disabled itself". Could this
> be the source of the problem?
>
> Part of the log is bellow. Does it help to work out why I lose my IP
> address?
>
> Thanks for your time.
>
> log -----
>
> --deleted
It could be. A while ago I had a machine on my network which
automatically installed zeroconf during an upgrade. Everything broke and
normal service resumed once I deleted it. It looks like the problem is
that both avahi-daemon and dhcp are trying to manage the eth0 interface.
My next guess, if you don't want to remove avahi etc, is to look at the
conf files and see if you can exclude eth0 from its control. I just had
a quick look at the avahi packages and there is one called avahi-autoipd
which is a daemon which automatically configures local IP addresses.
This may be the problem. Also, you have avahi-daemon running which is
something to do with auto configuration. I don't have either of these
installed (or zeroconf).

Looking at the log it would appear that dhcp managed to update your IP
address once so I don't think the problem lies there.


Good luck.

--
Keith Blow

Well, after playing around with it this weekend I found that it looks
like its NetworkManager that is causing the problem. If I shut down that
process my network like stays up. No idea why that should cause the problem.

I shall try and work out what cause the problem with the NetworkManager.
If I can't I will install it.

The other two problems still remain. If I lose my ISP connection I lose
my internal network connection and can't access windows machines from
windows machines using samba. Any ideas where to start look for those
two problems? I have looked in the logs for samba and do not get any
error messages (or any message) when I try to go from one windows
machine to the other. However, I do have a message in the log that say
it cannot create a user.

Thanks.

Keith Blow wrote:
> AAW wrote:
>> I did an uninstall and a reinstall but no change. So I have had a look
>> at the logs. I can see that it gets its IP address from my ISP :
>>
>> Nov 15 15:46:04 skamis dhclient: bound to 213.66.199.118 -- renewal in
>> 530 seconds.
>>
>> and then I can see that it loses its address :
>>
>> Nov 15 15:46:44 skamis dhclient: DHCPRELEASE on eth0 to 213.66.199.1
>> port 67
>>
>> ...
>>
>> Iautoip: Sending probe #0 for IP address 169.254.81.0.
>>
>>
>> but I don't really follow why. One thing I note in the log is the bit
>> about "Avahi detected that your currently configured local DNS server
>> serves a domain .local" leading to "Avahi disabled itself". Could
>> this be the source of the problem?
>>
>> Part of the log is bellow. Does it help to work out why I lose my IP
>> address?
>>
>> Thanks for your time.
>>
>> log -----
>>
>> --deleted
> It could be. A while ago I had a machine on my network which
> automatically installed zeroconf during an upgrade. Everything broke and
> normal service resumed once I deleted it. It looks like the problem is
> that both avahi-daemon and dhcp are trying to manage the eth0 interface.
> My next guess, if you don't want to remove avahi etc, is to look at the
> conf files and see if you can exclude eth0 from its control. I just had
> a quick look at the avahi packages and there is one called avahi-autoipd
> which is a daemon which automatically configures local IP addresses.
> This may be the problem. Also, you have avahi-daemon running which is
> something to do with auto configuration. I don't have either of these
> installed (or zeroconf).
>
> Looking at the log it would appear that dhcp managed to update your IP
> address once so I don't think the problem lies there.
>
>
> Good luck.
>

On Nov 19, 8:39 am, AAW <andrew.wall...@psy.umu.se_remove> wrote:
> Well, after playing around with it this weekend I found that it looks
> like its NetworkManager that is causing the problem. If I shut down that
> process my network like stays up. No idea why that should cause the problem.

NetworkManager should never be used on a server. A server should have
all it's interfaces have static adresses (except if the interfacs to
your ISP uses DHCP to get IP from the ISP, obviously)

NetworkManager is designed for desktop and laptops with one network
interface.

> I shall try and work out what cause the problem with the NetworkManager.
> If I can't I will install it.

You should uninstall NetworkManager from your servers.

> The other two problems still remain. If I lose my ISP connection I lose
> my internal network connection and can't access windows machines from
> windows machines using samba. Any ideas where to start look for those
> two problems?

Could be that avahi try to controll your ISP-interface, which it
shouldn't. Avahi is part of Apples Zeroconf, which makes your
computers announce themself on your local net under the name
yourhost.local, where yourhost is your computers name.
It can also set up a linklocal network addres (it's something like
169.??.0.0/16 I think), if you want to communicate with another
computer without setting up IP etc. If you use ethernet over
Firewire, it's good to have too, I guess.
Quite nice, on a laptop. Not on a server.
I you distribute your OGG and MP3-files with mt-daapd, your media
player on that net will be able to play those songs automagically. I
would configure avahi to only talk to your local network interface.
Avahi is nice to have on a local net, but not on your ISP net.

I would guess that NetworkManager and not setting static IP address on
your servers local network interface is the problem. NetworkManager
looks into /etc/network/interfaces to see what interfaces it can't
use. Read the documentation for NetworlManager in /usr/local/doc/
netw* something.

Good luck