[Ntp] Re: informal discussion of ptp security approach and mechanisms

[Doug Arnold <[e-mail]>]

Security is hard
Multicast security is really hard
PTP security with on-path support is pretty much impossible

But since much of the worlds critical infrastructure depends on PTP to work, some smart people need to figure something out to make PTP less vulnerable.

Here are some things that people have discussed:

  *
Leave the correction field out of the ICV and check it with heuristics (As proposed in NTP over PTP)
  *
Use a group key scheme where members of the group are authenticated on check in (Like GDOI or NTS for PTP)
  *
Get time from multiple timeTransmitters with as much path diversity as possible (like NTP does)
  *
Have Ethernet switches refuse to grant access to the PTP multicast addresses unless a device can produce a digital certificate that is on the authorized list (See IEEE 802.1X)

The IEEE 1588 working group purposely came to the NTP working group looking for help.  Since NTP works over the public internet you folks had to solve security already.  Also, in general there is more security expertise in the IETF membership than elsewhere.


  *
Doug

________________________________
From: Hal Murray <[e-mail]>
Sent: Sunday, July 7, 2024 4:47 PM
To: [e-mail] <[e-mail]>
Cc: Hal Murray <[e-mail]>
Subject: [Ntp] Re: informal discussion of ptp security approach and mechanisms


I've missed a key step in the big picture.

The basic idea in PTP is to modify a packet in flight.
One goal of security is to detect modified packets.

Those seem like conflicting ideas.  So what does "security" mean in the
context of PTP?


--
These are my opinions.  I hate spam.



_______________________________________________
ntp mailing list -- [e-mail]
To unsubscribe send an email to [e-mail]