Cisco Support Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Community Member

Odd behaviour with multiple ip helper-address's

Hi all,

I have a couple of 3550's running c3550-ipservicesk9-mz.122-25.SED which are displaying odd DHCP relay behaviour when multiple ip helper-address's are defined on any VLAN. For example:

interface Vlan4

ip address 192.168.4.1 255.255.255.224

ip helper-address 192.168.2.11

ip helper-address 192.168.2.12

no ip redirects

no ip unreachables

no ip proxy-arp

end

In this configuration, the DHCP unicast packets (relayed by the switch) only reache 192.168.2.11 and NOT 192.168.2.12. The behaviour is incontary to how my understanding of how multiple ip helper-address's should function - I expect to receive a the unicast packets (sourcing from the switch) at both DHCP servers. Extensive sniffing has confirmed that only the *1st* host defined in an ip helper-address statement receives the relayed packets.

I can confirm that relayed packets can reach both DHCP servers if I define only one ip helper-address under the VLAN interface, i.e.

interface Vlan4

ip address 192.168.4.1 255.255.255.224

ip helper-address 192.168.2.11

no ip redirects

no ip unreachables

no ip proxy-arp

end

OR:

interface Vlan4

ip address 192.168.4.1 255.255.255.224

ip helper-address 192.168.2.12

no ip redirects

no ip unreachables

no ip proxy-arp

end

I've spent hours pulling my hair out over this very strange problem. Please can someone assist?

Kind regards,

Tom Ranson

9 REPLIES

Re: Odd behaviour with multiple ip helper-address's

Tom,

I believe the behavior is if the first helper responds, it won't roll over to the next. The second one is used for failover.

HTH,

John

HTH, John *** Please rate all useful posts ***
Hall of Fame Super Blue

Re: Odd behaviour with multiple ip helper-address's

John

Sorry to have to contradict you but the L3 device should send unicast packets to all ip helper-addresses at the same time. The first to respond is usually the one used by the client.

Jon

Community Member

Re: Odd behaviour with multiple ip helper-address's

Hi Jon,

I agree - this is my understanding of how this should function.

Are you able to advise me as to why the behaviour I am seeing may be occuring?

Kind regards,

Tom

Hall of Fame Super Blue

Re: Odd behaviour with multiple ip helper-address's

Tom

"Are you able to advise me as to why the behaviour I am seeing may be occuring?"

Unfortunately no. Could be a bug perhaps. Have you tried another version of software to see if you get the same issue.

Perhaps if you posted the config something might jump out altho it's unlikely.

Jon

Re: Odd behaviour with multiple ip helper-address's

Jon,

This is how we learn :)

rated.

John

HTH, John *** Please rate all useful posts ***
Community Member

Re: Odd behaviour with multiple ip helper-address's

Hi John,

Many thanks for your prompt reply.

This goes against what Cisco state to be expected behaviour. Can you please point me towards any documentation to support this theory?

Kind regards,

Tom

Hall of Fame Super Gold

Re: Odd behaviour with multiple ip helper-address's

Tom

I am finding lots of references in Cisco documentation that specify that you can have multiple helper addresses. But I have not yet found one that specifies the expected result. But I certainly agree with Jon that the expected result is forwarding to each specified helper.

Since you seem to have a repeatable behavior I would suggest that you open a case with Cisco TAC about this.

HTH

Rick

Hall of Fame Super Gold

Re: Odd behaviour with multiple ip helper-address's

Tom

I now have found Cisco documentation that does describe the expected behavior of multiple helper addresses:

Cisco routers forward the DHCPDISCOVER message to all the helper addresses mentioned for that interface.

here is the link:

http://www.cisco.com/en/US/tech/tk648/tk36/technologies_tech_note09186a00800f0804.shtml

HTH

Rick

Community Member

Re: Odd behaviour with multiple ip helper-address's

Hi Rick,

Many thanks for your input - I'm positive that I'm not going mad now. It looks like I may have a bug on my hands; I'll add another '1' to my long list of open TAC cases :-/

Cheers,

Tom

579
Views
5
Helpful
9
Replies
CreatePlease to create content