Report information
The Basics
Id:
38144
Status:
rejected
Estimated:
16 hours (960 minutes)
Left:
16 hours (960 minutes)
Priority:
Low/Low
Queue:

BugTracker
Version Fixed:
(no value)
Version Found:
(no value)
Versions Affected:
(no value)
Versions Planned:
4.3.6
Priority:
P2 Normal
Severity:
S2 Normal
CVSS Score:
(no value)
CVE ID:
(no value)
Component:
(no value)
Area:
(no value)

Attachments
Dates
Created:Thu, 18 Dec 2014 07:31:31 -0500
Updated:Fri, 07 Jul 2017 20:00:43 -0400
Closed:Tue, 02 May 2017 15:39:04 -0400



This bug tracker is no longer active.

Please go to our Gitlab to submit issues (both feature requests and bug reports) for active projects maintained by Internet Systems Consortium (ISC).

Due to security and confidentiality requirements, full access is limited to the primary maintainers.

Subject: 'dhclient -6 -S -D LLT' should store the DUID-LLT
Date: Thu, 18 Dec 2014 13:31:25 +0100
To: dhcp-bugs@isc.org
From: Jiri Popelka <jpopelka@redhat.com>
Hello, by default, DHCPv6 dhclient creates DUID-LL if it is running in stateless mode (-6 -S) and doesn't write it into leases file, most likely because the DUID-LL is always generated the same. It's however possible to specify DUID to be of type LLT instead of LL with '-D LLT'. Rfc 3315 says that: 'Clients and servers using this type of DUID MUST store the DUID-LLT in stable storage.' That's not fulfiled in this case (-6 -S -D LLT), because it's generated each time again. It's not a big deal because the server doesn't store any info about 'stateless' clients, so it doesn't matter that the DUID-LLT is different each time. But there's a TAHI test which tests this, i.e. that DUID-LLT is still the same even in stateless mode. It's a test DHCP_CONF.7.1.9, part B. https://www.ipv6ready.org/docs/Phase2_DHCPv6_Conformance_Latest.pdf With regards, Jiri Popelka Red Hat, inc.

Message body is not shown because sender requested not to inline it.

On Thu Dec 18 12:31:31 2014, jpopelka@redhat.com wrote: > Hello, > > by default, DHCPv6 dhclient creates DUID-LL > if it is running in stateless mode (-6 -S) and > doesn't write it into leases file, most likely > because the DUID-LL is always generated the same. > > It's however possible to specify DUID to be of type LLT instead of LL > with '-D LLT'. Rfc 3315 says that: > 'Clients and servers using this type of DUID MUST > store the DUID-LLT in stable storage.' > That's not fulfiled in this case (-6 -S -D LLT), > because it's generated each time again. > > It's not a big deal because the server doesn't store any > info about 'stateless' clients, so it doesn't matter > that the DUID-LLT is different each time. > But there's a TAHI test which tests this, i.e. that > DUID-LLT is still the same even in stateless mode. > It's a test DHCP_CONF.7.1.9, part B. > https://www.ipv6ready.org/docs/Phase2_DHCPv6_Conformance_Latest.pdf => very funny: one needs a test suite (note I like the TAHI test :-) to discover such a problem. I don't know if the code has the provision to save and get back the DUID-LLT in stateless mode: if it is the case it will be easy to fix. If it is not, perhaps it will be easier to clarify the RFC 3315 which 1- is currently under revision 2- was written before stateless (and prefix delegation)... Thanks anyway and merry Christmas!
Hello Jiri: Given our resource constraints I cannot see any real world benefit to altering the behavior to satisfy a single TAHI test, especially when the test's relevancy is questionable. Unless you have strenuous objections we would prefer to simply close this one as "won't fix". As always thanks for you continued support and contributions. Regards, Thomas Markwalder