Internet-Draft scion-research_I-D July 2024
Meynell, et al. Expires 22 January 2025 [Page]
Workgroup:
PANRG
Internet-Draft:
draft-meynell-panrg-scion-research-questions-latest
Published:
Intended Status:
Informational
Expires:
Authors:
K. Meynell
SCION Association
J. García Pardo
ETH Zürich
T. Zäschke
ETH Zürich
N. Rustignoli
SCION Association

SCION Research Questions

Abstract

TODO Abstract here

About This Document

This note is to be removed before publishing as an RFC.

The latest revision of this draft can be found at https://scionassociation.github.io/scion-research_I-D/draft-meynell-panrg-scion-research-questions.html. Status information for this document may be found at https://datatracker.ietf.org/doc/draft-meynell-panrg-scion-research-questions/.

Discussion of this document takes place on the WG Working Group mailing list (mailto:panrg@irtf.org), which is archived at https://datatracker.ietf.org/rg/panrg. Subscribe at https://www.ietf.org/mailman/listinfo/panrg/.

Source for this draft and an issue tracker can be found at https://github.com/scionassociation/scion-research_I-D.

Status of This Memo

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.

Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."

This Internet-Draft will expire on 22 January 2025.

Table of Contents

1. Introduction

SCION is an inter-domain network architecture. Its core components, as deployed by some of its early adopters, are specified in [I-D.dekater-scion-dataplane], [I-D.dekater-scion-pki], [I-D.dekater-scion-controlplane] which are currently under ISE review.

The goal of this draft is to explore how SCION and its early deployments try to address open research questions in [RFC9217]. Specifically, there are still many open areas of research around path-aware networking, where SCION with its early deployment experiences and research efforts can provide a contribution. This can also be a starting point for discussions around long-term protocol evolution.

This draft assumes the reader is familiar with some of the fundamental concepts defined in the components specification.

Note: This is the very first version of the SCION research questions draft, and it merely contains a skeleton of potential topics to be further discussed in this draft. Any feedback is welcome and much appreciated. Thanks!

2. Conventions and Definitions

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.

3. Discovery, Distribution, and Trustworthiness of Path Properties

3.1. ISD, AS identity

The SCION protocol specifies 16 bits and 48 bits to identify the ISD and AS respectively. This identification is used at the data-plane level, in every packet to fully address the sender and receiver, and at the control-plane level, to identify the PCB sender and hops.

Whilst 48 bits for the AS will accommodate up to 2.81475e14 assignments which is likely to be more than sufficient for the future, using 16 bits for the ISD only offers 65,536 possible assignments. Further investigation on whether this is sufficient is required.

The following questions arise: (not comprehensive)

  • How many ASes do we expect in the SCION network model?

  • Can one AS belong to many ISDs?

  • Are AS numbers unique themselves, or only unique in combination with an ISD?

  • How many ISDs do we expect?

  • What is the ontology of an ISD?

    • Per geographic area?

    • Per legal jurisdiction?

    • Per capacity tier?

    • Per "type"? (meaning: any grouping that makes sense to their members)

    • Possible combinations of any previous "types"?

  • Note that, to achieve global connectivity, ISDs require core links to other ISDs. This reduces the number of ISDs to those that have core ASes that can directly connect to core ASes in other ISDs. The number of ISDs is still super-exponential asymptotically.

3.2. Beacon Selection Policies

Path discovery between SCION ASes relies on Path Construction Beacons (PCBs). In core beaconing PCBs are propagated omnidirectionally, unless this would cause a loop or exceed the maximum path length. A core AS selects a small number of paths to/from other core ASes, based on its beacon selection policy. It then propagates valid and policy compliant paths to neighbor ASes. The number of propagated beacons is limited to the best PCBs set size, in order to avoid that the number of paths (and beacons) grows exponentially with core network size.

Some potential questions are: * Limiting the number of beacons to a best PCBs set size per AS results in a partial view of the network being available to endpoints. To what point this affects endpoint's path-selection possibilities? * what is a good, practical, general purpose policy, that can fulfill conflicting requirements of both operators and endpoints, as highlighted in section 2.7 of [RFC9217]? * What are the desirable path properties (e.g. diversity, PCB Expiration Time, how recently the same PCBs was forwarded before).

3.3. DNS Service Binding (SVCB)

The DNS Service Binding [RFC9460], Section 14.3 allows a dedicated SCION Service Parameter to be specified.

Service Parameters allow the specification of alternative IP addresses or other parameters (such as ISD/AS numbers) for a given URL. This would be more elegant than using DNS TXT records.

Example of current entry:

$ dig +short ethz.ch TXT  | grep scion
"x-sciondiscovery=8041"
"scion=64-2:0:9,129.132.230.98"

With SVCB this may look like this for https:

dig +short https ethz.ch
1 . alpn="h2" scion=64-2:0:9,129.132.230.98

SVCB is also planned to be supported by Happy Eyeballs v3 [I-D.draft-pauly-v6ops-happy-eyeballs-v3-01].

3.4. Segment Dissemination

A path look-up in SCION works analogous to a DNS query (section 4 of [I-D.dekater-scion-controlplane]):

  • The source endpoint queries the control service in its own AS.

  • The local AS has already at least one segment to one core AS of its local ISD. It uses it to query the core AS' control service for segments to core ASes of the remote ISD.

  • The local AS has now segments also from its local ISD core ASes to core ASes in the remote ISD. The local AS queries the remote ISD's core ASes for segments to the destination AS.

  • The local AS returns all these segments to the endpoint, to be combined.

Control services may return a large number of path segments for some queries. This can cost considerable bandwidth while at the same time overloading clients with an unnecessarily large numbers of segments.

  • This problem may be more acute in ASes with many endpoints (e.g. IoT), or for resource-constrained endpoints.

  • Getting a full path to a remote endpoint may require three queries to control services.

There are multiple possible and independent solution steps here:

  • Predefine some policies that can be resolved by the control plane, e.g. ANY, BEST_LATENCY, BEST_BANDWIDTH, BEST_PRICE, BEST_CO2. For these, the control plane could simply calculate 5-10 compliant paths and return these. Moreover these could be cached for commonly requested remote ASes. If a user requires a custom policy they can still resort to requesting actual segments.

  • Caching of paths. An open question concerns when a control service should return a chached path, versus performing a recursive path lookup.

An example including the number of core segments between different ISDs as of 2024-07-12 in the SCION production network is shown in Table 1.

Table 1: core segment count examples
src ISD dst ISD segments returned
64 64 337
64 65 240
64 67 60

3.5. Periodic beacon propagation

The SCION control plane protocol specifies that beacons should be propagated periodically. Is this really necessary?

  • For path freshness, only the initial AS emitting the PCB needs to originate beacons periodically, and others can disseminate immediately.

  • As response to link failures or availability of new paths, beacon services can respond instantly.

If no periodic propagation is necessary for path freshness or to respond to link failures, the periodic propagation would only be used for the discovery of new paths at each interval, enhancing the scalability and path diversity.

3.6. Beacon optimization and extensibility

Communication requirements vary according to source, destination, and application. Satisfying all these requirements either requires discovering all paths in the network, or optimizing the creation of paths during the beaconing process. Selecting the 5 shortest paths per destination for each beaconing period may not satisfy all requirements that different applications, on different endpoints, and on different ASes, will have. The beacon selection process, the criteria and metrics that they carry, and the adaptability of them all have a strong impact in the traffic engineering of the individual ASes, and of the inter-domain communication as a whole. See question 2.7 of [RFC9217].

  • What optimization functions should be applied to beacons and what metrics should be considered when propagating them ? Is the set of properties composed of path length, peering ASes, path disjointness, PCB last reception, and path lifetime enough?

  • How do we extend the metrics with new dimensions, such as bandwidth, latency, geo-position, etc?

  • Who should select these functions?

  • How should the outcome of these functions be verified?

  • How can multiple functions be applied concurrently, for different source and destination applications?

  • How should end-ASes express their desired requirements to the inter-domain control plane?

  • How do these requirements translate into concrete optimization functions?

  • How would standardization of the functions look like?

  • The functions changing over time:

    • How can optimization functions adapt to incorporate these changes?

    • How to achieve fast adaptation of optimization functions?

  • See also: IREC [TABAEIAGHDAEI2023]

3.7. DRKey

DRKey is a key distribution system that scales well with the number of endpoints in the network. It relies on two things:

  • Two sides of a key: a fast side, and a slow side. Sometimes called fast and slow side of the derivation. The ability of deriving a key very quickly on the fast side is necessary for most of its use cases.

  • A grouping of endpoints (such as ASes): The pieces necessary to derive a key, namely the L1 keys, are communicated to each keystore at each grouping (e.g., a keystore per AS).

The questions related to DRKey are the following ones (not comprehensive):

  • Do we want to have any possible authorization that is at the moment carried out at the data-plane to be moved to the control-plane? This could include e.g. authorization to deliver traffic depending on the source, but also information like port numbers/ranges per source, etc.

    • Could this obsolete firewalls? What else would be necessary?

    • What do we mean when we say "authorize transit"?

  • Could perfect forward secrecy DRKey be useful, and should we research it?

    • What would be the trust model? Do end-users trust their ASes to ephemerally create personal keys?

    • What would be the attacker model?

    • Which use cases are relevant?

For more info: [I-D.garciapardo-panrg-drkey].

3.8. SCMP Authentication

In SCION, endpoints are responsible to select alternate paths in case of failure. One approach to detect failures is to rely on signals from the network, such as SCMP (SCION Control Message Protocol) messages. These signals must be authenticated in order to be trusted by endpoints. This reflects a question raised in section 2.7 of [RFC9217].

One option is to use DRKey as the mechanism to use to derive the authentication key, where the fast path would be on the infrastructure side (e.g. the border router in the case of an interface down type of message), and the slow side being on the intended endpoint destination for that SCMP message (e.g. the endpoint receiving the SCMP interface down message). Another option is to leverage the control-plane PKI.

However, we have identified a number of possible issues (not comprehensive):

  • Denial of Service/Capability Attacks: If an endpoint receives (too) many SCMP messages, it will need (too) many resources just to authenticate their origin. Most of these messages could just be sent to the endpoint to exhaust its processing capacity.

  • Sending an SCMP message in certain cases might be an amplification factor: If a border router sends an SCMP message (e.g. interface down) on all cases, even with small packets, there is the risk of having that border router sending a lot of traffic to a possibly unintended recipient, e.g. when the packet is not source validated. In addition, validation may trigger additional requests for keys.

3.10. NAT

Currently, the SCION implementation is not compatible out-of-the-box with NAT'ed devices, regardless of whether these devices are end-hosts, or even running SCION services. This is due to the (UDP-IP) underlay being modified by the NAT mechanism, but not the internal destination SCION address. Although this does not concern the SCION protocols themselves, we want to check that this will not be a problem. Critically, the SCION header needs to contain the SRC address as seen by the border router so that the border router can forward incoming response packets to the correct NAT device and port.

Possible solutions:

  • With IPv6 underlay, this problem disappears. // TODO Clarify why it disappears? Is the idea that we can remove NATs if everyone would use IPv6?

  • Introduce a mechanism so that the SCION border router can report the NATed address to an endpoint (similar to a STUN server).

4. Dataplane stability

4.2. Reverse Path Refreshment

When a client and server communicate, return traffic should usually follow the same reverse path. If the server uses that path for a long period of time, the path will eventually expire. How should the server determine the path for return traffic and at which layer? How to avoid this being a vector pf path hijackings?

There are some relevant points for the discussion:

  • In order to send data back to the client, the server needs to store the path locally (analogous to storing the client's 4-tuple in an TCP/IP scenario).

  • More generally, if multiple paths are used to contact the server, which one of those would be used to reply? Should this be responsibility of the transport protocol, as in the case of QUIC-MP [I-D.ietf-quic-multipath]?

  • How long before path expiration should the client and server still use a path? The client can choose from paths that will not expire in a short period of time, but it does not control for how long the server will attempt to use it (i.e. how long it will take the server to send the complete response).

4.2.1. Proposed Solutions (not comprehensive)

  • The server MUST ask the control plane for a path, regardless of the client's policy.

  • The client SHOULD (somehow) send a new packet with a new path, prompting the server to use this path from now on.

  • The client and server agree, via a path policy specification, on which kinds of paths are okay for the server to use. This solution implies a standard specification of this path policy.

  • Leave the solution to the application and transport layers. Transport protocols may require keep alive messages, or already support multiple paths. Applications should know for how long they are willing to read a response from a server. With this knowledge these two layers can easily determine when to send a new path (analogous to connection migration in QUIC [RFC9000]), so that the server is instructed to use it for the next replies.

  • The server must ask the control plane for a path, regardless of the client's policy.

  • The client (somehow) sends a new packet with a new path, prompting the server to use this path from now on.

There are some nuances: Usually the server's API will store the initial address of the client to be used through all the session.

A related question: how long before expiration should a path be used?

Is reverse path refresh a relevant problem?

  • Contra: It is probably rare that a server needs to send data for a long time without the application layer protocol requiring the client to ever answer back.

  • Pro: The client may happen to have an old-ish path. If we can't refresh, the client always needs to consider whether a path is valid "long enough", which might only be possible to guess.

  • Contra: Sending keep-alives sounds like a connection based protocol. It also means we need to figure out when to stop sending keep-alives.

  • Contra: It may be better to solve this in the application layer or in the overlay protocol, where we we know more about potential length of the session, whether this is a singular request/answer type of exchange, or whether more frequent keep-alives are required anyway.

5. Interfaces for Path Awareness

6. Implications of Path Awareness for the Transport and Application Layers

This question relates to 2.5 in [RFC9217].

7. Naming

To be discussed

8. Security Considerations

TODO Security

9. IANA Considerations

This document has no IANA actions.

10. References

10.1. Normative References

[I-D.dekater-scion-controlplane]
de Kater, C., Rustignoli, N., and S. Hitz, "SCION Control Plane", Work in Progress, Internet-Draft, draft-dekater-scion-controlplane-04, , <https://datatracker.ietf.org/doc/html/draft-dekater-scion-controlplane-04>.
[I-D.dekater-scion-dataplane]
de Kater, C., Rustignoli, N., and S. Hitz, "SCION Data Plane", Work in Progress, Internet-Draft, draft-dekater-scion-dataplane-02, , <https://datatracker.ietf.org/doc/html/draft-dekater-scion-dataplane-02>.
[I-D.dekater-scion-pki]
de Kater, C., Rustignoli, N., and S. Hitz, "SCION Control-Plane PKI", Work in Progress, Internet-Draft, draft-dekater-scion-pki-06, , <https://datatracker.ietf.org/doc/html/draft-dekater-scion-pki-06>.
[I-D.ietf-quic-multipath]
Liu, Y., Ma, Y., De Coninck, Q., Bonaventure, O., Huitema, C., and M. Kühlewind, "Multipath Extension for QUIC", Work in Progress, Internet-Draft, draft-ietf-quic-multipath-10, , <https://datatracker.ietf.org/doc/html/draft-ietf-quic-multipath-10>.
[RFC2119]
Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, , <https://www.rfc-editor.org/rfc/rfc2119>.
[RFC8174]
Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, , <https://www.rfc-editor.org/rfc/rfc8174>.
[RFC9000]
Iyengar, J., Ed. and M. Thomson, Ed., "QUIC: A UDP-Based Multiplexed and Secure Transport", RFC 9000, DOI 10.17487/RFC9000, , <https://www.rfc-editor.org/rfc/rfc9000>.
[RFC9217]
Trammell, B., "Current Open Questions in Path-Aware Networking", RFC 9217, DOI 10.17487/RFC9217, , <https://www.rfc-editor.org/rfc/rfc9217>.
[TAPS]
IETF, "Transport Services Working Group", , <https://datatracker.ietf.org/wg/taps>.

10.2. Informative References

[I-D.draft-pauly-v6ops-happy-eyeballs-v3-01]
Pauly, T., Schinazi, D., Jaju, N., and K. Ishibashi, "Happy Eyeballs Version 3: Better Connectivity Using Concurrency", Work in Progress, Internet-Draft, draft-pauly-v6ops-happy-eyeballs-v3-01, , <https://datatracker.ietf.org/doc/html/draft-pauly-v6ops-happy-eyeballs-v3-01>.
[I-D.garciapardo-panrg-drkey]
de los Galanes, J. A., Krähenbühl, C., Rothenberger, B., and A. Perrig, "Dynamically Recreatable Keys", Work in Progress, Internet-Draft, draft-garciapardo-panrg-drkey-03, , <https://datatracker.ietf.org/doc/html/draft-garciapardo-panrg-drkey-03>.
[KRAHENBUHL2023]
Krähenbühl, C., Wyss, M., Basin, D., Lenders, V., Perrig, A., and M. Strohmeier, "FABRID: Flexible Attestation-Based Routing for Inter-Domain Networks", , <https://www.usenix.org/conference/usenixsecurity23/presentation/krahenbuhl>.
[LEGNER2020]
Legner, M., Klenze, T., Wyss, M., Sprenger, C., and A. Perrig, "EPIC: Every Packet Is Checked in the Data Plane of a Path-Aware Internet", , <https://netsec.ethz.ch/publications/papers/Legner_Usenix2020_EPIC.pdf>.
[RFC9460]
Schwartz, B., Bishop, M., and E. Nygren, "Service Binding and Parameter Specification via the DNS (SVCB and HTTPS Resource Records)", RFC 9460, DOI 10.17487/RFC9460, , <https://www.rfc-editor.org/rfc/rfc9460>.
[TABAEIAGHDAEI2023]
Tabaeiaghdaei, S., Wyss, M., Giuliari, G., van Bommel, J., Zehmakan, A. N., and A. Perrig, "Inter-domain Routing with Extensible Criteria", , <https://netsec.ethz.ch/publications/papers/IREC_arXiv.pdf>.

Acknowledgments

TODO acknowledge.

Authors' Addresses

Kevin Meynell
SCION Association
Juan A. García Pardo Giménez de los Galanes
ETH Zürich
Tilmann Zäschke
ETH Zürich
Nicola Rustignoli
SCION Association