mirror of
https://github.com/ntop/n2n.git
synced 2024-10-18 15:42:49 +05:30
f577d997a7
- Legacy features are now moved to the legacy directory with readme - Keyschedule feature is not supported anymore
34 lines
1.5 KiB
Markdown
34 lines
1.5 KiB
Markdown
# Removed Features
|
|
|
|
This folder contains a list N2N legacy features which have been dropped due to
|
|
maintainance cost versus effective use and benefits.
|
|
|
|
Multiple Transops
|
|
-----------------
|
|
|
|
N2N used to initialize all the available transops and use the "tick" function of
|
|
the transops to decide which transop to use before sending a packet. This however
|
|
has the following problems:
|
|
|
|
- It only works with the keyfile, whereas with normal encryption we inizialize and
|
|
keep structures that we don't need.
|
|
- It is unfeasable as an edge node is required to implement all the transops in order
|
|
to properly talk with other edge nodes (via keyfile).
|
|
- It rises the complexity of the code.
|
|
- It is not clear which transop will be used.
|
|
- Mixing multiple encyptions together is not necessarily a good idea to improve security
|
|
as a vulnerability in at least one encryption method will leak some information.
|
|
|
|
Keyfile and Key Rotation
|
|
------------------------
|
|
|
|
The keyfile mechanism allowed N2N users to specify a keyfile to be used to periodically
|
|
rotate keys and encryption methods. However, it has the following problems:
|
|
|
|
- This feature is obscure for most of the users and poorly documented.
|
|
- It is tightly integrated in the core whereas it is used by only a few people (if any).
|
|
|
|
In conclusion the main problem is the complexity that it adds to the code. In a possible
|
|
future rework this could be integrated as an extention (e.g. a specific trasop) without
|
|
rising the core complexity.
|