Skip to main content

5. Impact of Proposed Changes

This section discusses the impact of the proposed changes on legacy devices, datagram generation in updated devices, middleboxes, and header compression.

5.1. Impact on Legacy Internet Devices​

Legacy uses of the IPv4 ID field consist of fragment generation, fragment reassembly, duplicate datagram detection, and "other" uses.

Current devices already generate ID values that are reused within the source address/destination address/protocol tuple in less than the current estimated Internet MDL of two minutes. They assume that the MDL over their end-to-end path is much lower.

Existing devices have been known to generate non-varying IDs for atomic datagrams for nearly a decade, notably some cellphones. Such constant ID values are the reason for their support as an optimization of ROHC [RFC5225]. This is discussed further in Section 5.4. Generation of IPv4 datagrams with constant (zero) IDs is also described as part of the IP/ICMP translation standard [RFC6145].

Many current devices support fragmentation that ignores the IPv4 Don't Fragment (DF) bit. Such devices already transit traffic from sources that reuse the ID. If fragments of different datagrams reusing the same ID (within the source address/destination address/protocol tuple) arrive at the destination interleaved, fragmentation would fail and traffic would be dropped. Either such interleaving is uncommon or traffic from such devices is not widely traversing these DF-ignoring devices, because significant occurrence of reassembly errors has not been reported. DF-ignoring devices do not comply with existing standards, and it is not feasible to update the standards to allow them as compliant.

The ID field has been envisioned for use in duplicate detection, as discussed in Section 4.1. Although this document now allows IPv4 ID reuse for atomic datagrams, such reuse is already common (as noted above). Protocol accelerators are known to implement IPv4 duplicate detection, but such devices are also known to violate other Internet standards to achieve higher end-to-end performance. These devices would already exhibit erroneous drops for this current traffic, and this has not been reported.

There are other potential uses of the ID field, such as for diagnostic purposes. Such uses already need to accommodate atomic datagrams with reused ID fields. There are no reports of such uses having problems with current datagrams that reuse IDs.

Thus, as a result of previous requirements, this document recommends that IPv4 duplicate detection and diagnostic mechanisms apply IPv6-compatible methods, i.e., methods that do not rely on the ID field (e.g., as suggested in [RFC6621]). This is a consequence of using the ID field only for reassembly, as well as the known hazard of existing devices already reusing the ID field.

5.2. Impact on Datagram Generation​

The following is a summary of the recommendations that are the result of the previous changes to the IPv4 ID field specification.

Because atomic datagrams can use arbitrary IPv4 ID values, the ID field no longer imposes a performance impact in those cases. However, the performance impact remains for non-atomic datagrams. As a result:

Sources of non-atomic IPv4 datagrams MUST rate-limit their output to comply with the ID uniqueness requirements. Such sources include, in particular, DNS over UDP [RFC2671].

Because there is no strict definition of the MDL, reassembly hazards exist regardless of the IPv4 ID reuse interval or the reassembly timeout. As a result:

Higher-layer protocols SHOULD verify the integrity of IPv4 datagrams, e.g., using a checksum or hash that can detect reassembly errors (the UDP and TCP checksums are weak in this regard, but better than nothing).

Additional integrity checks can be employed using tunnels, as supported by the Subnetwork Encapsulation and Adaptation Layer (SEAL) [RFC5320], IPsec [RFC4301], or the Stream Control Transmission Protocol (SCTP) [RFC4960]. Such checks can avoid the reassembly hazards that can occur when using UDP and TCP checksums [RFC4963] or when using partial checksums as in UDP-Lite [RFC3828]. Because such integrity checks can avoid the impact of reassembly errors:

Sources of non-atomic IPv4 datagrams using strong integrity checks MAY reuse the ID within intervals that are smaller than typical MDL values.

Note, however, that such frequent reuse can still result in corrupted reassembly and poor throughput, although it would not propagate reassembly errors to higher-layer protocols.

5.3. Impact on Middleboxes​

Middleboxes include rewriting devices such as network address translators (NATs), network address/port translators (NAPTs), and other address-sharing mechanisms (ASMs). They also include devices that inspect and filter datagrams but that are not routers, such as accelerators and firewalls.

The changes proposed in this document may not be implemented by middleboxes; however, these changes are more likely to make current middlebox behavior compliant than to affect the service provided by those devices.

5.3.1. Rewriting Middleboxes​

NATs and NAPTs rewrite IP fields, and tunnel ingresses (using IPv4 encapsulation) copy and modify some IPv4 fields; all are therefore considered datagram sources, as are any devices that rewrite any portion of the source address/destination address/protocol/ID tuple for any datagrams [RFC3022]. This is also true for other ASMs, including IPv4 Residual Deployment (4rd) [De11], IVI [RFC6219], and others in the "A+P" (address plus port) family [Bo11]. It is equally true for any other datagram-rewriting mechanism. As a result, they are subject to all the requirements of any datagram source, as has been noted.

NATs/ASMs/rewriters present a particularly challenging situation for fragmentation. Because they overwrite portions of the reassembly tuple in both directions, they can destroy tuple uniqueness and result in a reassembly hazard. Whenever IPv4 source address, destination address, or protocol fields are modified, a NAT/ASM/rewriter needs to ensure that the ID field is generated appropriately, rather than simply copied from the incoming datagram.

Specifically:

Address-sharing or rewriting devices MUST ensure that the IPv4 ID field of datagrams whose addresses or protocols are translated comply with these requirements as if the datagram were sourced by that device.

This compliance means that the IPv4 ID field of non-atomic datagrams translated at a NAT/ASM/rewriter needs to obey the uniqueness requirements of any IPv4 datagram source. Unfortunately, translated fragments already violate that requirement, as they repeat an IPv4 ID within the MDL for a given source address/destination address/protocol tuple.

Such problems with transmitting fragments through NATs/ASMs/rewriters are already known; translation is typically based on the transport port number, which is present in only the first fragment anyway [RFC3022]. This document underscores the point that not only is reassembly (and possibly subsequent fragmentation) required for translation, it can be used to avoid issues with IPv4 ID uniqueness.

Note that NATs/ASMs already need to exercise special care when emitting datagrams on their public side, because merging datagrams from many sources onto a single outgoing source address can result in IPv4 ID collisions. This situation precedes this document and is not affected by it. It is exacerbated in large-scale, so-called "carrier grade" NATs [Pe11].

Tunnel ingresses act as sources for the outermost header, but tunnels act as routers for the inner headers (i.e., the datagram as arriving at the tunnel ingress). Ingresses can always fragment as originating sources of the outer header, because they control the uniqueness of that IPv4 ID field and the value of DF on the outer header independent of those values on the inner (arriving datagram) header.

5.3.2. Filtering Middleboxes​

Middleboxes also include devices that filter datagrams, such as network accelerators and firewalls. Some such devices reportedly feature datagram de-duplication that relies on IP ID uniqueness to identify duplicates, which has been discussed in Section 5.1.

5.4. Impact on Header Compression​

Header compression algorithms already accommodate various ways in which the IPv4 ID changes between sequential datagrams [RFC1144] [RFC2508] [RFC3545] [RFC5225]. Such algorithms currently assume that the IPv4 ID is preserved end-to-end. Some algorithms already allow the assumption that the ID does not change (e.g., ROHC [RFC5225]), where others include non-changing IDs via zero deltas (e.g., Enhanced Compressed RTP (ECRTP) [RFC3545]).

When compression assumes a changing ID as a default, having a non-changing ID can make compression less efficient. Such non-changing IDs have been described in various RFCs (e.g., footnote 21 of [RFC1144] and cRTP [RFC2508]). When compression can assume a non-changing IPv4 ID -- as with ROHC and ECRTP -- efficiency can be increased.

5.5. Impact of Network Reordering and Loss​

Tolerance to network reordering and loss is a key feature of the Internet architecture. Although most current IP networks avoid gratuitous such events, both reordering and loss can and do occur. Datagrams are already intended to be reordered or lost, and recovery from those errors (where supported) already occurs at the transport or higher protocol layers.

Reordering is typically associated with routing transients or where flows are split across multiple paths. Loss is typically associated with path congestion or link failure (partial or complete). The impact of such events is different for atomic and non-atomic datagrams and is discussed below. In summary, the recommendations of this document make the Internet more robust to reordering and loss by emphasizing the requirements of ID uniqueness for non-atomic datagrams and by more clearly indicating the impact of these requirements on both endpoints and datagram transit devices.

5.5.1. Atomic Datagrams Experiencing Reordering or Loss​

Reusing ID values does not affect atomic datagrams when the DF bit is correctly respected, because order restoration does not depend on the datagram header. TCP uses a transport header sequence number; in some other protocols, sequence is indicated and restored at the application layer.

When DF=1 is ignored, reordering or loss can cause fragments of different datagrams to be interleaved and thus incorrectly reassembled and discarded. Reuse of ID values in atomic datagrams, as permitted by this document, can result in higher datagram loss in such cases. Situations such as this already can exist because there are known devices that use a constant ID for atomic datagrams (some cellphones), and there are known devices that ignore DF=1, but high levels of corresponding loss have not been reported. The lack of such reports indicates either a lack of reordering or a loss in such cases or a tolerance to the resulting losses. If such issues are reported, it would be more productive to address non-compliant devices (that ignore DF=1), because it is impractical to define Internet specifications to tolerate devices that ignore those specifications. This is why this document emphasizes the need to honor DF=1, as well as that datagram transit devices need to retain the DF bit as received (i.e., rather than clear it).

5.5.2. Non-atomic Datagrams Experiencing Reordering or Loss​

Non-atomic datagrams rely on the uniqueness of the ID value to tolerate reordering of fragments, notably where fragments of different datagrams are interleaved as a result of such reordering. Fragment loss can result in reassembly of fragments from different origin datagrams, which is why ID reuse in non-atomic datagrams is based on datagram (fragment) maximum lifetime, not just expected reordering interleaving.

This document does not change the requirements for uniqueness of IDs in non-atomic datagrams and thus does not affect their tolerance to such reordering or loss. This document emphasizes the need for ID uniqueness for all datagram sources, including rewriting middleboxes; the need to rate-limit sources to ensure ID uniqueness; the need to not reuse the ID for retransmitted datagrams; and the need to use higher-layer integrity checks to prevent reassembly errors -- all of which result in a higher tolerance to reordering or loss events.