Bug 1390 – ICMPv6 Redirect are handled correctly only for /64 networks

Bug 1390 - ICMPv6 Redirect are handled correctly only for /64 networks


Summary:	ICMPv6 Redirect are handled correctly only for /64 networks

Status:	RESOLVED DUPLICATE of bug 1697

Product:	ns-3
Classification:	Unclassified
Component:	ipv6
Version:	ns-3-dev
Hardware:	All All

Importance:	P5 normal
Assigned To:	Tommaso Pecorella

URL:
Keywords:

Depends on:
Blocks:
	Show dependency tree / graph

Reported:	2012-03-11 20:12 UTC by Tommaso Pecorella
Modified:	2013-06-16 05:16 UTC (History)
CC List:	3 users (show)

See Also:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.

Description Tommaso Pecorella 2012-03-11 20:12:40 UTC

The bug is rather obscure, so using an example.

Let's use this network:
A-----B-----C

A's IPV6 address is 2001:db8:1::1/64
B's IPV6 addresses are  2001:db8:1::2/64 and 2001:db8:2::2/64
C's IPV6 address is 2001:db8:2::1/64

All goes fine.

However, if the IPv6 addresses are as follows:
A's IPV6 address is 2001:db8:cafe:f00d:1::1/80
B's IPV6 addresses are  2001:db8:cafe:f00d:1::2/80 and 2001:db8:cafe:f00d:2::2/80
C's IPV6 address is 2001:db8:cafe:f00d:2::1/80

THEN for each packet forwarded by node B, an ICMPv6 Redirect is sent (erroneously).

The solution to this issue is all but trivial. However now the issue is documented, thus it's a feature.

Comment 1 Tommaso Pecorella 2012-03-25 08:33:37 UTC

As a workaround, one can disable ICMPv6 Redirect using the Ipv6L3Protocol's attribute "SendIcmpv6Redirect" (added in changeset 7793 - 1c81a8a91b1e).

Leaving the bug open as the behavior is not the intended one anyway.

Comment 2 shiguowang 2013-05-16 03:59:40 UTC

Hi Tommaso,
I think This will happen because the prefix can only be 64 right now, you can reference the function of SetBase() of file src/internet/helper/ipv6-address-helper.cc


Best Regards
gwshi

Comment 3 Tommaso Pecorella 2013-05-17 07:05:59 UTC

Hi gwshi,

yep, it's my same idea. The issue involves aslo the RT behaviour tho. Without CIDR handling and/or DHCPv6 the actual behaviour isn't a big issue.
The bug, however is open as a future reference. It should be fixed but, right now, we miss also a case where it's strongly needed.

Cheers,

T.

Comment 4 Tommaso Pecorella 2013-06-16 05:16:27 UTC


*** This bug has been marked as a duplicate of bug 1697 ***