Problem with IP aliases

Problem with IP aliases

Post by Alessandro Fiorin » Fri, 19 Nov 1999 04:00:00



I have a problem with IP aliasing: I have an old server with a dozen of
IPs aliased on the same ethernet card which is connect to a router to
the external Internet. When I changed the server with a new one using
the same IP configuration, the aliased IP are unreachable from the
external network, traceroute stops at the router address, only the main
IP of the card works; but if I reconenect the old server all works
perfectly. After that I reconeccted the new one and after one hour 2 of
the aliased IPs started to be reachable from the external network, I
re-reconetted the old server and now the 2 IPs that started to work on
the new one became unreachable when used on the old. After another hour
with the old server connected this 2 IPs restarted to work flawlessy.
What's the problem ? I think the router doesn't correctly updates his
ARP cache for the aliased IPs. Is this assertion correct ? How can I
force it to notice the change in the network configuration (I can't
manipulate directly its configuration) ?
I'm using Linux 2.2.13 with RedHat 6.1.
TIA for every suggestion.
 
 
 

1. DNS problems with IP aliasing.

I ran into a problem last week when we converted an ISP site from a
Linux-based machine to SCO Release 5.  I had originally installed the
SCO box on an unused IP address in parallel with the Linux system
moving things over one at a time and adding IP aliases for web and ftp
servers as we went.  The final step was to take the Linux box offline
entirely and alias its old IP address to the NIC on the SCO system.

This caused problems with DNS queries for zone transfers and some of
the queries were totally ignored.  I had set up our system as a
secondary DNS server to their domain, but couldn't get and zone
transfers until I reconfigured their SCO box so the primary IP address
was the same as the original Linux system (and the same as the
published address for the primary name server by the NIC).  Normal
BIND queries for e-mail, ping, telnet, etc., worked generally although
there were reports that some web servers weren't responding.  It seems
that the udp side may have been working while the tcp functions used
for zone transfers didn't.

I don't know whether this problem is a quirk in the named program or
possibly something in the DNS protocols that prevent any responses
from anything but the definitive SOA address being recognized.

I'm posting this in the hopes that it may help someone else avoid this
particular pitfall in the future, and to see if there's some good
reason this happens or whether it's something that needs to be fixed.

Bill
--

UUCP:              camco!bill   2835 82nd Avenue S.E. S-100
FAX:           (206) 232-9186   Mercer Island, WA 98040; (206) 947-5591
http://www.celestial.com/
SPEED COSTS MONEY -- HOW FAST DO YOU WANT TO GO?

2. savecore !

3. A little problem with IP aliasing of user PPP

4. sound Blaster live setup problem !!!

5. Q: Problem with IP Aliasing

6. SiI 3112 sata_error

7. routing problem to ip alias

8. how do I get my SummaExpression to work ?

9. Problem assigning IP aliases for NIC's

10. Problem deleting IP aliases

11. problem with IP aliasing...

12. arp/routing problem with ip aliases in 2.2.13

13. Default route problem with IP aliased interface