DAD Ipv6 Duplicate Address Detection

Hi everyone,
My questions are; If a node starts a DAD process and send a NS, is it creates an entry on its neighbour cache for target address? Also if it recieve a NA packet for this NS we know that target address is duplicate and we can`t assign it, but what about neighbour cache? Is this NA update neighbor cache?
Consequently , can DAD process update neighbor cache and used for reachibilty?
All replies will be appreciated.
Gokmen...

Hi Gokmen,
During the TENTATIVE Step of Autoconfiguration we verify the IPv6 Addresses.
We start with Link-Local Address with DAD and if we receive a NA, two cases:
- Normal case, the interface is disabled for IPv6 (see remark below)
- SeND. We retry twice. If after 3 attempts, computed CGA are not unique we disable the Interface for IPv6 and report an Error (I tested it!)
Now about DAD itself, it is explained in rfc4862
http://tools.ietf.org/html/rfc4862#section-5.4.4
Quite funny the RFC says that the interface is disabled for a link-local only if it is derived from a MAC Address otherwise IP operation MAY continue... Quite funny as I don't see how we could continue with a duplicated Link-local address!!!!
Now when it is the turn of other Addresses, if we receive a NA, the Address is not initialized by the local Node.
AFAIK, while the interface is in TENTATIVE mode we are not supposed to build the Neighbor Cache but only verify our own addresses and other task very well defined.
The address is not initialized and may be once the interface is enabled and no longer in TENTATIVE mode we can resolve this Neighbor MAC Address but I would be surprised that it would be done while in TENTATIVE mode during DAD!!!
I stay tuned... As somebody may correct me if I am wrong!
You may be interested by page 16 and 17, click on the graphics to see the animations...
http://www.fredbovy.com/Tutorial/IPv6Tutorial-RELEASE2.html
For more details about autoconfiguration you may be interested by this video (Presentation also available from slideshare, please check my wiki page)
http://www.youtube.com/watch?v=958sRC306UE
It would be funny
Kind Regards,
Fred
15 years ccie #3013
18 years ccsi #33517 (former #95003)
IPv6 Forum Gold Certified Engineer
IPv6 Forum Gold Certified Trainer
Email: [email protected]
Web: http://www.fredbovy.com
Wicki: http://www.fredbovy.com/MediaWiki
Twitter: http://twitter.com/#!/Fr
Skype: FredericBovy

Similar Messages

Nexus 5548 - %ARP-3-DUP_SRCIP_PROBE: Duplicate address Detected

Hello.
Looking for some assistance on the following please, I'll try to include a lot of detail:
I noticed we're receiving repeating messages on a pair of Nexus 5548UP switches, regarding duplicate IP's, and the IP's in question are the mgmt0 of each respective Nexus switch; which are connected to each other and uplinked to two 3750X switches which are also linked to each other (not as a stack, with optics).
Here is a sample of the syslog message: %ARP-3-DUP_SRCIP_PROBE: arp [####] Duplicate address Detected. Probe packet received from zzzz.zzzz.zzzz on mgmt0 with destination set to our local ip, xxx.xxx.xxx.xxx
And in ascii art, here is a diagram: SW1&2 are 3750X (running 15.x code) and NX1&2 are Nexus 5548UP (running 6.x code)
SW1---SW2
| |
NX1---NX2
Pretty basic box/"ring", just in case the diagram doesn't format well, with rapid spanning-tree in effect. The Nexus are Layer2 only, and the 3750X's are Layer 2 & 3 /w HSRP on a few SVI's in them. All vlans can pass over the connecting trunks. And on the Nexus, the management interface (mgmt0) is connected to 3750X (NX1 to SW1 and NX2 to SW2); e.g. management is done over this vrf interface, versus it flowing through a trunk interface to a SVI on the Nexus.
I know these messages to be erroneous as far an being actual duplicate IP's on the network, and have seen a few posts that indicate a command ("no ip arp gratuitous hsrp duplicate") typically referencing a Nexus 7000 setup has helped, but obviously as described that's not involved here. Note: Placing that command on the mgmt0 interface of the two nexus had no positive effect. The messages didn't start showing until after HSRP was enabled on the applicable 3750X SVI's - so seems a likely correlation.
Essentially looking to clear this up (ideally) or suppress the messages since they're false. Any ideas?
Thanks!

For those encountering this same issue, here is some information that may lead you to a resolve.
I decided to open a TAC case to get some direct assistance, and I was directed to the 3750X's and a new feature in the 15.x code, "ip device tracking". I was directed to remove that without specific details / steps on how to best do it (e.g. "no ip device tracking" does not remove it from the global config, as of 15.2(1)E anyway and is the most current at the time of this posting), and that didn't take, giving the message: % IP device tracking is disabled at the interface level by removing the relevant configs. However, after doing some research on the command, I can across some similar posts on 15.x code and duplicate IP messages.... Here's two posts I referenced:
https://supportforums.cisco.com/thread/2244042
https://supportforums.cisco.com/thread/2239656
I opted for the "no macro auto monitor" command in the 3750X's in global config, as mentioned in the above post(s) - This did stop the messages from occurring in the Nexus switches. IMPORTANT NOTE: I noticed that it did appear to bounce all the switch ports however, which could impact production - Be advised!
I cannot speak to if the interface level command 'nmsp attach suppress' is a better route to go, which is also referenced in the posts (first one primarily) - I went the other route to not have all the extra config line entries. I have asked Cisco what the true impact of the "no macro auto monitor" command is, and am waiting a response - Ideally there's little to none, probably depending on if you want to make use of device tracking or not... in my case, I suspect I don't have a need for this and thus there's no impact - I shall see.
It would be ideal if code is released to disable this service / feature in the future more gracefully and/or code changes are done to remedy this issue so you can have the feature enabled (or not). I akin this much like the "vstack" service that was introduced around 12.2(55/58)SE IOS that couldn't be disabled & left an open TCP port (not good for security scans), then with 12.2(58)SE2 they allowed it to be disabled (e.g. "no vstack" in global config).
Hopefully these details prove of value.
Regards.
NOTE: Since it doesn't seem you can"answer" your own posts, which seems silly; please note I consider this item resolved / answered.

IPv6 Duplicate Address issue

I'm having a problem involving IPv6 with stateless autoconfig on my network. I have a Cisco 800 series router and we have several VLAN interfaces each configured with an IPv6 prefix. The problem I'm noticing is that whenever any host on the network tries to start up with stateless autoconfig they immediately detect a duplicate address for whatever address they are trying to use.
I performed a packet capture and what I'm seeing is that when the host selects an IPv6 address to use, it performs the Neighbor Solicitation to check if the address is in use. I immediately see the exact same Neighbor Solicitation message echo'd back with the source MAC being the Cisco router. This causes the host to reject the address as a duplicate since it is receiving a Neigbor Soliciation for the same address it is attempting to use.
This happens on all of the VLANs I have configured for IPv6. The basic VLAN config is like this:
interface Vlan109
description Engineering VLAN
ip address .....
ip helper-address .....
ip nat inside
ip virtual-reassembly in
ip tcp adjust-mss 1300
ipv6 address HE-ENG ::/64 eui-64
ipv6 enable
ipv6 nd ra interval 60
Attached is a PCAP with some ICMPv6 traffic. You can see that the NS messages are duplicated and the source of the duplicates seems to be the Cisco router (70:ca:9b:e0:94:2e). Does anyone know what is going on here?

Hi,
Under you interface val config can you try adding:-
int vlan 109
ipv6 address autoconfig
Worth a try
Regards
Alex

WLC Duplicate IP address detected for AP-Manager Interface

I am getting an error log in the WLC saying, its IP address is duplicate by another machine with MAC address A.B.C.D
But this MAC address A.B.C.D is the MAC address of the AP-Manager Interface in the same controller.
Model No.                   AIR-WLC2106-K9
Software Version                 7.0.116.0
%LWAPP-3-DUP_IP: spam_lrad.c:27626 Adding client 58:b0:35:83:72:86 to exclusion list due to IP Address conflict with AP 'AP_DUXO_3'
%LWAPP-3-DUP_AP_IP: spam_lrad.c:27612 Duplicate IP address detected for AP AP_DUXO_3, IP address of AP 10.184.1.224, this is a duplicate of IP on another machine (MAC address 58:b0:35:83:72:86)
Cisco AP Identifier.............................. 1
Cisco AP Name.................................... AP_DUXO_3
Country code..................................... US - United States
Regulatory Domain allowed by Country............. 802.11bg:-A     802.11a:-A
AP Country code.................................. US - United States
AP Regulatory Domain............................. 802.11bg:-A    802.11a:-N
Switch Port Number .............................. 1
MAC Address...................................... cc:ef:48:1a:e4:af
IP Address Configuration......................... Static IP assigned
IP Address....................................... 10.184.1.224
IP NetMask....................................... 255.255.0.0
Gateway IP Addr.................................. 10.184.20.2
Domain...........................................
Name Server......................................
NAT External IP Address.......................... None
CAPWAP Path MTU.................................. 1485
Telnet State..................................... Enabled
Ssh State........................................ Disabled
Cisco AP Location................................ DUXO_BOX
Cisco AP Group Name.............................. default-group
Does anyone have an issue like this ?

Are you sure this MAC address 58:b0:35:83:72:86 isn't some type of Apple device? Its OUI is registered to apple. How do clients get ip addresses DHCP? It appears that the IP 10.184.1.224 is statically assigned to your ap-manager and that this client 58:b0:35:83:72:86 is either getting that same IP from DHCP or the client is statically assigning it themselves.

Duplicate IP address detected in WCS

Hi to Cisco Support Community,
I would like to verify some questions regarding our network setup for wireless devices.
We have here in our client a 5508 WLC version 6.0.196.0 and WCS version 7.0.164.0 using windows platform.
My concern is that there is always a duplicate IP address detected by the WCS and when we trace the machine given its MAC address, we found out that the machine is a printer with a static IP of different IP address not the same to the management IP address of the WLC.
I have already disabled the port where the machine detected is terminated and still experiencing the same issue.
Is this a software bug or some sort of network issue? Please advice.
Thank you in advance!
Regards,
Joghson

Hi,
I have raised the issue with TAC and they told me that this is a software issue, we are currently making packet sniffs which is needed by the software developer. I will give you updates on this.
Thanks!

IPv6 Duplicate error msg on GNS3

Hello,
I am trying IPv6 Lab on GNS3, but I am again and again getting the following duplicate error message on my GNS3 router,
%IPV6-4-DUPLICATE: Duplicate address FE80::C000:13FF:FE80:0 on FastEthernet0/0
I tried changing my Interface Link-Local address, but no result, Please let me know how to get it resolved ?

Hi,
Unfortunately, you have not solved the problem - you have only deactivated the mechanism that reports the problem. The issue with the packets sent by your router looping back to the same router persists, and it may cause another problems in future. I caution you against running your setup like this, especially if you use it for learning - because of the packets being fed back to the router, the behavior of your topology may be different.
Just curious: when you run your router attached to the VirtualBox instance for some time (more than one minute) and reports the IPv6 address conflict (i.e. without the workaround you've suggested earlier), what does the show cdp neighbor show? Does this input contain an entry about this very router? If yes, it definitely confirms the packets being reflected back to the router for some obscure reason.
Do you perhaps have an option of running your GNS3 on another machine, or on another operating system? At this point, I am fairly sure that the problem is somewhere in the way GNS3 ties together a VirtualBox instance and an IOS router instance. However, this is not about solving IOS issues anymore (and it never was as it turns out), this is about possible issue in GNS3 or in your particular setup. Unfortunately, there are many variables in play, including the precise settings of GNS3, VirtualBox, your particular operating system and the software running on it.
Best regards,
Peter

How can I turn off IPv6 temporary addresses in a enterprise environment

So in a default configuration Vista and Windows 7 clients will use IPv6 temporary address (per RFC 3041), but I would like to be able to disable this with a GPO.
I know I can do this by using a startup script tied to a GPO using the netsh interface ipv6 set privacy state=disabled store=persistent but I really do not want to run a logon script especially when as you can see in the command it is a persistent
setting.
Any ideas on using a registry based GPO for this?

Hi,
OK, I understand what you want. But after I use the Process Monitor to capture the behavior, I found that a lot of registry keys would be changed. So you can not set it via registry unless the IPv6 is disabled.
Thanks for understanding.
Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread. ”
Thank you again for the answer.
As a follow up question... How does one suggest/request a change in default behavior in a MS product?
While the RFC 3041 addresses might be nice to have for the average consumer they are not ideal in a enterprise environment. I have found that the temporary addresses will register in DNS, but this doesn't completely solve the issue of tracking and accountability,
you would still need a application to query, correlate and store that information.
So it would be nice if the business versions of Windows Vista/7/2K8 and the next client OS would not have this behavior by default.

How to get rid of duplicate addresses all over the place?

Duplicate addresses appears all over the places: on my mac (in the mail program) and in my ios devices.
I suspect that it occurs during some synchronisation activities but how to get rid of this once and for all?
Thanks for your help. It helps also if some tell me that I am not alone in that kind of trouble

iCloud- Resolving duplicate Contacts after setting up iCloud Contacts

How to create a duplicate address book contact?

This seems so simple yet I cannot determine how to create a new duplicate address book contact. I have a few contacts from the same company and want to change the variable data instead of re entering the data
I can create a new contact, copy the data of the record I want to duplicate yet when I paste the info into the new record - address book puts everything into one field.
Why not a duplicate contact command?
GRRRR

Rich,
According to Address Book Help:
To duplicate an address card:
Select an address card.
Choose Edit > Copy, and then choose Edit > Paste.
Works for me

[svn:bz-trunk] 21394: bug fix for watson 2887837 Not getting duplicate session detected error when same flex client id is used from two different HTTP sessions in CRX .

Revision: 21394
Revision: 21394
Author:   [email protected]
Date:     2011-06-16 12:34:13 -0700 (Thu, 16 Jun 2011)
Log Message:
bug fix for watson 2887837 Not getting duplicate session detected error when same flex client id is used from two different HTTP sessions in CRX.
get the sessions id before we invalidate the duplicate session.
Checkintests pass
Modified Paths:
    blazeds/trunk/modules/core/src/flex/messaging/endpoints/BaseHTTPEndpoint.java

For our profect I think this issue was caused as follows:
Believing that remoting was full asynchronous we fired a 2 or 3 remote calls to the server at the same time ( within the same function ) - usually when the users goes to a new section of the app.
This seemed to trigger the duplicate http session error since according to http://blogs.adobe.com/lin/2011/05/duplication-session-error.html two remote calls arriving before a session is created will cause 2 sessions to be created.
Our current solution ( too early to say it works ) is to daisy chain the multiple calls together .
Also there seemed to be an issue where mobile apps that never quit ( thanks Apple! ) caused the error when activated after a few hours.
I guess the session expires on the server and the error above occurs on activation.
So the mobile apps now ping the server with a remote call when activated after sleeping for more than one hour.
All duplicate http errors are silently caught and reported.
Fingers crossed we won't get any more!

Can't set ipv6 static address

As the topic states, I can not setup a static IP to save my life. I've enabled ipv6, I have a hostname6.ce1 I put the hostname (madrid-ipv6) in the hostname6.ce1 file but it will not stay set.
I've tried the addif command, and it creates a logical interface, but I can't ping out with the address that is set.
I'm sure I am missing something, and I am new to Solaris. I've got ipv6 static addresses setup in linux with no problems. It's just Solaris 10 that is kicking my butt.
If anyone can point me in the right direction I would be in your debt.
Thanks!

edit: To darren: I've tried that as well. When the system boots, it gives me the error: "Failed to assign ipv6 address to ce1" or something similar to that.
hmmm, I've tried that, it still doesn't seem to be working. Here is my hostname6.ce1 file:
madrid-ipv6
addif fd01::e2/64 up
this is the output of ifconfig -a6:
bash-3.00# ifconfig -a6
lo0: flags=2002000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv6,VIRTUAL> mtu 8252 index 1
inet6 ::1/128
ce0: flags=2004841<UP,RUNNING,MULTICAST,DHCP,IPv6> mtu 1500 index 2
inet6 fe80::203:baff:fe9b:92d/10
ether 0:3:ba:9b:9:2d
ce1: flags=2000840<RUNNING,MULTICAST,IPv6> mtu 1500 index 3
inet6 fe80::203:baff:fe9b:92d/10
ether 0:3:ba:9b:9:2d
ce1:1: flags=2000841<UP,RUNNING,MULTICAST,IPv6> mtu 1500 index 3
inet6 fd01::e2/64
and this is what I get when I ping fd01::f2:
bash-3.00# ping fd01::f2
ICMPv6 Address Unreachable from gateway fd01::e2
for icmp6 from fd01::e2 to fd01::f2
ICMPv6 Address Unreachable from gateway fd01::e2
for icmp6 from fd01::e2 to fd01::f2
ICMPv6 Address Unreachable from gateway fd01::e2
for icmp6 from fd01::e2 to fd01::f2
I've got to be missing something. This is all on a private switch that doesn't need to connect to the net. I need to connect an iscsi VTL to this sun system but the first step is to be able to ping the damn thing. I've checked the networking to make sure everything is connected. I can ping the VTL using IPV4, but not IPV6. I've checked on a linux system and it can ping the VTL over IPV6 so I know it is setup properly. This is just driving me nuts.
Edited by: jonwheeler on Sep 11, 2008 9:47 AM

After security updates on Windows 2008 R2 ping returns ipv6 ip addresses

After installing the updates shown on screenshot attached the server running windows 2008 r2 returns IPv6 ip addresses when pinging any device. Also Public Folder got populated with many icons that were not there. This is a server running terminal and
rdp services. Please help me identify which update may be causing this issue

This one may help.
How to disable IPv6 or its components in Windows
or alternately you can
ping -4 {ipaddress}
Regards, Dave Patrick ....
Microsoft Certified Professional
Microsoft MVP [Windows]
Disclaimer: This posting is provided "AS IS" with no warranties or guarantees , and confers no rights.

Dimension processing error: "duplicate id detected"

Hi All,
Im facing an issue where when i try to process a dimension, i get an error message ."error while processing dimensions"
"duplicate id detected" .The reason for this is there are two identical id's with in the same dimension. "pu" & "PU".
Im not able to root cause, how BPC allowed processing of these identical dimensions in the first place. But becoz of this, whene ever i try to add a new dimension member and process the same, i get the above mentioned error.
Have you ever encountered this error in your applications.
If so, what is the way out?
Also, Is there a solution ,which does not involve deleting & re-creating the dimensions.
Thanks in advance,
sridhar

Hi Sridhar,
Best way is delete both the line and process it. The wild fact in BPC is that if if you have used ABC as dimension member in cost center, you can not have same name as dimension member in profit center as well.
Is it not wild? So best way is delete the dimension member..but if you have posted some transaction data u have to delete the transaction data first..

Duplicate addresses

Here's a weird one. Virtually all the records in my address book recently disappeared - this is a periodic problem with Address Book. The records return when Sync asks if I want to delete more than a thousand records on .Mac. I select cancel and the records miraculously reappear in the Address Book.
But this week, something different happened. When the addresses reappeared, over a thousand had been modified by duplication. Every work and home address is duplicated. What could have caused this bloating and how can I get rid of the duplicates? I took a backup of the database with duplicates from another of my Macs and just rebuilt the Address Book from that, yet it still contains the duplicate addresses. (Incidentally, it's only the addresses that have been duplicated in each record - not the names, phone numbers and emails.)
One possible cause is that I recently purchased a fourth Mac and when I asked to Sync it warned that there were two computers with my username. I chose a different name to deal with that. But on one of my original Macs Retrospect is now telling me that there are two "indistinguishable" hard drives with the same username. I can only see one but there could be some sort of ghost in my system.
And suggestions appreciated. Cheers.

Sync invited me to choose another name. I just entered a variant of my usual user name (firstnamelastname), putting my middle initial in. That was on the new Mac which is running OS 10.5.something. On the older Mac, running OS 10.4.11, I've also changed the account name although, of course, the shortname remains the same. Retrospect reports the problem of the disk with the same name on the old Mac, but not on the new one.
Another interesting thing: only Address Book entries with addresses have been updated. Address Book entries with just names and phone numbers or emails have not been updated (according to the light grey "Last updated" text in the bottom right of the window. It's as though just one or two fields in the underlying database have been duplicated (Home address and Work address).

2504 WebAuth and IPv6 RADIUS Accounting (IPv6-Framed-Address)

Hi Board,
I'm playing around with RADIUS Accounting in combination with local web authentication on the wireless LAN controller.
So far so good - everything works well, but I'm missing the "IPv6-Framed-Address" in the RADIUS accounting messages.
The only thing I can see is the v4 framed IP address and the "Framed-IPv6-Prefix". According to the configuration guide
http://www.cisco.com/c/en/us/td/docs/wireless/controller/8-0/configuration-guide/b_cg80/b_cg80_chapter_0101001.html#ID807
the "IPv6-Framed-Address" should be sent by the WLC. I took a capture on a span port of the WLC to verify this. Anybody else experiencing this behavior or is it a simple misconfiguration on my side? In the client details I can see the global IPv6 addresses and the link-local.
I tested it on a WLC 2504 with 8.0.100.0 code.
Cheers
Johannes

Hi Board,
I'm playing around with RADIUS Accounting in combination with local web authentication on the wireless LAN controller.
So far so good - everything works well, but I'm missing the "IPv6-Framed-Address" in the RADIUS accounting messages.
The only thing I can see is the v4 framed IP address and the "Framed-IPv6-Prefix". According to the configuration guide
http://www.cisco.com/c/en/us/td/docs/wireless/controller/8-0/configuration-guide/b_cg80/b_cg80_chapter_0101001.html#ID807
the "IPv6-Framed-Address" should be sent by the WLC. I took a capture on a span port of the WLC to verify this. Anybody else experiencing this behavior or is it a simple misconfiguration on my side? In the client details I can see the global IPv6 addresses and the link-local.
I tested it on a WLC 2504 with 8.0.100.0 code.
Cheers
Johannes

DAD Ipv6 Duplicate Address Detection

Similar Messages

Maybe you are looking for