same here. I'm using iproute-2.4.7-14. I think it's a kernel 
networking problem. I'm using kernel-smp-2.6.6-1.435.2.3 and I have:

[root@hostname root]# route
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    
Use Iface
XXX.XXX.XXX.XX  *               255.255.255.XXX U     0      0        
0 eth0
169.254.0.0     *               255.255.0.0     U     0      0        
0 eth0
default         xxxxxx-xxx-xx.x 0.0.0.0         UG    0      0        
0 eth0


I'm expecting something like

[root@hostname2 root]# route
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    
Use Iface
XXX.XXX.XXX.XX  *               255.255.255.XXX U     0      0        
0 eth0
127.0.0.0       *               255.0.0.0       U     0      0        
0 lo
169.254.0.0     *               255.255.0.0     U     0      0        
0 eth0
default         xxxxxx-xxx-xx.x 0.0.0.0         UG    0      0        
0 eth0


Although the system seems to be fine without it, I've tried using tin 
(1.6.2 and 1.7.5) with stunnel (> 
http://www.redhat.com/archives/redhat-list/2002-July/msg01483.html).

Although I'm able to read postings, I'm unable to post any message 
successfully. I get the following message after quitting 
tin: "Warning Can't connect to server -- Article will be spooled."

If I didn't set up stunnel successfully I would have gott the 
message: 
tin 1.7.5 release 20040615 ("Gighay") [UNIX] (c) Copyright 1991-2004 
Iain Lea.
Connecting to 127.0.0.1:5432...
Connection to 127.0.0.1: Connection refused
Giving up...
Failed to connect to NNTP server 127.0.0.1. Exiting...

However, I was able to connect to the secure NNTP server successfully 
but unable to post successfully. I think this has to do with the 
missing lo entry (loopback entry).

I tried "route add -net 127.0.0.0" but got the message:
SIOCADDRT: Invalid argument

Not sure about the correct syntax but according to the man page of 
route, it assumes that the "lo" device was previously set up 
correctly with ifconfig. Any idea how to manually add the normal 
loopback entry or set up the "lo" device correctly with ifconfig?
           

ok. now I tried the following:

"route add -host 127.0.0.1 lo"    followed by
"route add -net 127.0.0.0 netmask 255.255.0.0 lo"

I get the following in my routing table:

localhost.local *               255.255.255.255 UH    0      0        
0 lo
127.0.0.0       *               255.255.0.0     U     0      0        
0 lo


unfortunately that does not solve my problem. I could read but still 
not post with tin and stunnel:
Reading config file...
Warning Can't connect to server -- Article will be spooled.

sorry, it should be:

route add -net 127.0.0.0 netmask 255.0.0.0 lo  (not 255.255.0.0)


only the above is needed.


I've tried downgrading to older kernel releases but it doesn't help.
Maybe there's some incompatibilty between tin and stunnel with Fedora
Core 2.

if I'm not wrong, the following don't work with (tin + stunnel):

kernel-smp-2.6.5-1.358
kernel-smp-2.6.6-1.435.2.3
kernel-smp-2.6.7-1.494.2.2

kernel-2.6.5-1.358
kernel-2.6.6-1.435.2.3
kernel-2.6.7-1.494.2.2

maybe the problem lies with the stunnel I'm using:

stunnel-4.05-1

When installing the default fedora core 2 the lo device is listed in 
the routing tables, but after doing an up2date -u the lo device no 
longer shows up in the routing table

Bug also find with "ip route" command. Seems to me more like kernel
bug. I will keep on testing. 

reassigned to stunnel

OK now I'm worried.  I have this problem on Enterprise Linux 3 on two
systems, different h/w platforms.  I noticed the problem after the
"update 3" went on them, but can't be certain it didn't happen before
that.  Should I raise a bug against RHEL 3 or is this one enough?
For certain RHEL 3 off the CD does NOT have this problem.

Ian Laurie, which bug are you referring to? the stunnel or missing lo 
entry bug? I also have RHEL3 Update 3 running but there is no missing 
lo entry and no stunnel bug problem. What I'm using specifically:

cat /etc/redhat-release
Red Hat Enterprise Linux ES release 3 (Taroon Update 3)

Running Taroon Update 3... all updates applied.

I refer to the bug of this thread, missing lo entry.  I do not know if
this is a stunnel issue or not.

I have the issue on two different systems, different CPUs (well P IV
and P IV H/T, but they are different kernels as the latter is running
the SMP version).

I "noticed" the issue just after update 3, but as I said it could have
happened before.  However the lo entries were NOT missing when I
originally reported this bug against FC2 because at the time I used
the RHEL 3 boxes as a reference and they DID have the lo entry then. 
That was only a month ago.

BTW - both systems have a FULL CD set install on them (check box at
the bottom which says EVERYTHING)... dunno if that is a clue, I may
have more stuff on my system than you do.

I'm seeing this with an up to date FC2 running 2.6.8-1.521.  As of yet
no obvious problems caused by the routing table omission.  Despite it
not being listed everything seems to get routed to lo on 127.0.0.0/8.

Fedora Core 2 is now maintained by the Fedora Legacy project for
security updates only. If this problem is a security issue, please
reopen and reassign to the Fedora Legacy product. If it is not a
security issue and hasn't been resolved in the current FC3 updates or
in the FC4 test release, reopen and change the version to match.

$ /sbin/ip addr ls
1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 brd 127.255.255.255 scope host lo
...

127/8 is treated as a local host IP address, a routing table entry is
not necessary.