Re: Troubleshooting connection loss (continued)
- From: Bit Twister <BitTwister@xxxxxxxxxxxxxxxx>
- Date: Thu, 08 Nov 2007 20:21:06 GMT
On Thu, 08 Nov 2007 19:08:10 GMT, Allen Weiner wrote:
(I posted a similar thread to this newsgroup on October 29. Due to
continuing problems, I'm opening this thread.)
I run Fedora 7 and use Verizon DSL. My modem is a Westell 6100-E90
modem/router.I have no other networking hardware. My DSL connection
usually runs well, but about once every seven to ten days I lose my
Internet connection. I can regain my connection by rebooting Fedora.
I've not been able to regain my Internet connection without a reboot
(e.g. "service network restart" hangs).
GKrellm: The eth0 monitor shows zero activity. When I have an Internet
connection, the eth0 monitor shows continuous activity, even when I'm
not accessing anything.
You have activity like arp ack/req, email checks, ntp time
check/sync.... Not to mention anything sent by the router
"even when you'r not accessing anything"
Nov 8 11:51:44 localhost kernel: NETDEV WATCHDOG: eth0: transmit timed out
No experience with netdev watchdog. I wonder if it is tearing down
your connection.
[root@localhost ~]# route -n
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use
Iface
Well, there goes your routing so the "no route to 192.168.1.1" is correct.
[root@localhost ~]# ifconfig
Yep, and that is sad.
Following the connection loss, but prior to issuing "service network
restart", I issued "route -n". The output was the same as before the
connection loss.
What you posted above, did not show us that fact.
Double check if /avahi/ needs to be disabled on boot
avahi-daemon 0:off 1:off 2:off 3:on 4:on 5:off 6:off
avahi-daemon is still not disabled.
You need to click up a terminal and do the following:
su - root
service avahi-daemon stop
chkconfig --del avahi-daemon
========== head -15 /etc/hosts ===========
192.168.1.1 gateway
That bites, you should have a local host entry.
I suggest one for the node, if you would give it a node name.
Looking in your dhcp lease file we find
option domain-name "myhome.westell.com";
So you could name it that if you like, but I wonder what is going on
because I see
$ host westell.com
westell.com has address 216.203.29.175
westell.com mail is handled by 10 cluster9.us.messagelabs.com.
westell.com mail is handled by 20 cluster9a.us.messagelabs.com.
Something is not looking good in the router, from what I can guess so far.
You better get into the router and veify that the dns servers belong
to verizon.
Come on just to make eveything standard, Please modify /etc/sysconfig/network
to set a unique node name, not localhost.localdomain
/etc/sysconfig/network
NETWORKING=yes
NEEDHOSTNAME=no <==== add this line
HOSTNAME=darkstar.home.invalid <==== pick your own node name
put .invalid on the end
Now your /etc/hosts should have something like:
127.0.0.1 localhost
192.168.1.1 gateway
192.168.1.47 darkstar.home.invalid darkstar
I will suggest changing your ip address to 192.168.1.140
It is hard to prove if you are running static or dhcp.
Change ip addy in /etc/sysconfig/network-scripts/ifcfg-eth0,
/etc/hosts to match ip addy and reboot. You should have no problem.
======== cat /etc/sysconfig/network-scripts/ifcfg-eth0 ==========
# Intel Corporation 82557/8/9 [Ethernet Pro 100]
DEVICE=eth0
ONBOOT=yes
BOOTPROTO=none
HWADDR=00:07:e9:01:b2:09
TYPE=Ethernet
USERCTL=yes
IPV6INIT=no
PEERDNS=yes
NETMASK=255.255.255.0
IPADDR=192.168.1.47
GATEWAY=192.168.1.1
BOOTPROTO= seems to indicate static but look,
renew 3 2007/11/7 05:23:24;
rebind 3 2007/11/7 15:31:25;
expire 3 2007/11/7 18:31:25;
Why are you getting a new dhcp lease??????
Something is still not right. Maybe PEERDNS=yes caused the lease request.
When you made your changes, did you use the gui interface, or did you
just edit files?
If edited files for eth0, use gui interface to modify eth0.
======== tail -18 /var/lib/dhclient/dhclient-eth0.leases ==========
rebind 3 2007/11/7 12:23:43;
expire 3 2007/11/7 15:23:43;
}
lease {
interface "eth0";
fixed-address 192.168.1.47;
option subnet-mask 255.255.255.0;
option routers 192.168.1.1;
option dhcp-lease-time 86400;
option dhcp-message-type 5;
option domain-name-servers 192.168.1.1,192.168.1.1;
option dhcp-server-identifier 192.168.1.1;
option broadcast-address 255.255.255.255;
option domain-name "myhome.westell.com";
renew 3 2007/11/7 05:23:24;
rebind 3 2007/11/7 15:31:25;
expire 3 2007/11/7 18:31:25;
What troubleshooting step should I do next?
After my suggested changes,
echo "nameserver 192.168.1.1" > /etc/resolv.conf
cp /dev/null /var/lib/dhclient/dhclient-eth0.leases
reboot, because of node name changes, and when the system comes up, do a
cat /var/lib/dhclient/dhclient-eth0.leases
and verify no dhcp lease.
If so, I would
cd /etc/sysconfig/network-scripts/
cp ifcfg-eth0 ifcfg-eth0.bkup
change
BOOTPROTO=static
PEERDNS=no
in /etc/sysconfig/network-scripts/ifcfg-eth0
cp /dev/null /var/lib/dhclient/dhclient-eth0.leases
service network restart
echo "nameserver 192.168.1.1" > /etc/resolv.conf
cat /var/lib/dhclient/dhclient-eth0.leases
I would now think there would be a null dhclient-eth0.leases file.
if restart fails
cp ifcfg-eth0.bkup ifcfg-eth0
and change
PEERDNS=no
cp /dev/null /var/lib/dhclient/dhclient-eth0.leases
echo "nameserver 192.168.1.1" > /etc/resolv.conf
service network restart
cat /var/lib/dhclient/dhclient-eth0.leases
cat /etc/resolv.conf should show only one nameserver line. If two,
your dhcp client is still running.
Do veify ip address is 192.168.1.140 in
ifconfig
.
- Follow-Ups:
- Re: Troubleshooting connection loss (continued)
- From: Allen Weiner
- Re: Troubleshooting connection loss (continued)
- From: Allen Weiner
- Re: Troubleshooting connection loss (continued)
- References:
- Troubleshooting connection loss (continued)
- From: Allen Weiner
- Troubleshooting connection loss (continued)
- Prev by Date: Re: Machines on LAN
- Next by Date: How to change TTL of Linux
- Previous by thread: Troubleshooting connection loss (continued)
- Next by thread: Re: Troubleshooting connection loss (continued)
- Index(es):
Relevant Pages
|
