18 KiB
| title | name | category | status | tags | editor | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| RELIABLE-CHANNEL-API | Reliable Channel API definition | Standards Track | raw |
|
Logos Messaging Team |
Table of contents
- Table of contents
- Abstract
- Motivation
- Syntax
- API design
- Components
- Procedures
- The Reliable Channel API
- Security/Privacy Considerations
- Copyright
Abstract
This document specifies the Reliable Channel API,
an application-level interface that sits between the application layer and the MESSAGING-API plus P2P-RELIABILITY, i.e., application <-> reliable-channel-api <-> messaging-api/p2p-reliability.
It bundles segmentation, end-to-end reliability via Scalable Data Sync (SDS), rate limit management, and a pluggable encryption hook into a single interface for sending and receiving messages reliably.
Motivation
The MESSAGING-API provides peer-to-peer reliability via P2P-RELIABILITY, but does not provide high end-to-end delivery guarantees from sender to recipient.
This API addresses that gap by introducing:
- SEGMENTATION-API to handle large messages exceeding network size limits.
- SDS to provide causal-history-based end-to-end acknowledgement and retransmission.
- Rate Limit Manager to comply with RLN constraints when sending segmented messages.
- Encryption Hook to allow upper layers to provide a pluggable encryption mechanism. The main motivation for encryption is to protect the payload and content_topic from MessageEnvelope, defined in MESSAGING-API.
The separation between Reliable Channels and encryption ensures the API remains agnostic to identity and key management concerns, which are handled by higher layers.
Syntax
The keywords "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC2119.
API design
Architectural position
The Reliable Channel API sits between the application layer and the Messaging API, as follows:
┌────────────────────────────────────────────────────────────┐
│ Application Layer │
└───────────────────────────┬────────────────────────────────┘
│
┌───────────────────────────▼────────────────────────────────┐
│ Reliable Channel API │
│ ┌──────────────┐ ┌─────┐ ┌───────────────┐ ┌──────────┐ │
│ │ Segmentation │ │ SDS │ │ Rate Limit Mgr│ │Encryption│ │
│ │ │ │ │ │ │ │ Hook │ │
│ └──────────────┘ └─────┘ └───────────────┘ └──────────┘ │
└───────────────────────────┬────────────────────────────────┘
│
┌───────────────────────────▼────────────────────────────────┐
│ Messaging API │
│ (P2P Reliability, Relay, Filter, Lightpush, Store) │
└────────────────────────────────────────────────────────────┘
IDL
A custom Interface Definition Language (IDL) in YAML is used, consistent with MESSAGING-API.
Components
Segmentation
See SEGMENTATION-API.
Scalable Data Sync (SDS)
SDS provides end-to-end delivery guarantees using causal history tracking.
- Each sent chunk is registered in an outgoing buffer.
- The recipient sends acknowledgements back to the sender upon receiving chunks.
- The sender removes acknowledged chunks from the outgoing buffer.
- Unacknowledged chunks are retransmitted after
acknowledgementTimeoutMs. - SDS state MUST be persisted using the
persistencebackend configured inSdsConfig.
Rate Limit Manager
The Rate Limit Manager ensures compliance with RLN rate constraints.
- It tracks how many messages have been sent in the current epoch (only the first chunk of each message counts toward the rate limit; subsequent chunks are exempt).
- When the limit is approached, chunk dispatch MUST be delayed to the next epoch.
- The epoch size MUST match the
epochSizeMsconfigured inRateLimitConfig.
Encryption Hook
The Encryption Hook provides a pluggable interface for upper layers to inject encryption.
- The hook is optional; when not provided, messages are sent unencrypted.
- Encryption is applied per chunk, after segmentation and SDS registration.
- Decryption is applied per chunk, before SDS delivery.
- The
IEncryptioninterface MUST be implemented by the caller. - The Reliable Channel API MUST NOT impose any specific encryption scheme.
Procedures
Outgoing message processing
When send is called with a ReliableEnvelope whose channelId is non-empty, the implementation MUST process envelope.messageEnvelope.payload in the following order:
- Segment: Split the payload into chunks as defined in SEGMENTATION-API.
- Apply SDS: Register each chunk with the SDS layer to track acknowledgements and enable retransmission.
- Encrypt: If an
IEncryptionimplementation is provided, encrypt each chunk before transmission. - Rate Limit: If
RateLimitConfig.enabledistrue, delay dispatch as needed to comply with RLN epoch constraints. - Dispatch: Send each chunk via the underlying MESSAGING-API.
Incoming message processing
When a chunk is received from the network, the implementation MUST process it in the following order:
- Decrypt: If an
IEncryptionimplementation is provided, decrypt the chunk. - Apply SDS: Deliver the chunk to the SDS layer, which emits acknowledgements and detects gaps.
- Reassemble: Once all chunks for a message have been received, reassemble and emit a
reliable:message:receivedevent.
Retransmission
A certain Reliable Channel MUST retransmit unacknowledged chunks after acknowledgementTimeoutMs,
according to the SDS channel state,
up to maxRetransmissions times.
If a chunk is not acknowledged after all retransmission attempts, a reliable:message:send-error event MUST be emitted.
Rate limiting
When RateLimitConfig.enabled is true, the implementation MUST space chunk transmissions
to comply with the RLN epoch constraints defined in epochSizeMs.
Chunks MUST NOT be sent at a rate that would violate the RLN message rate limit for the active epoch.
Encryption
The encryption field in ReliableChannelConfig is intentionally optional.
The Reliable Channel API is agnostic to encryption mechanisms.
When an IEncryption implementation is provided, it MUST be applied as described in Outgoing message processing and Incoming message processing.
The Reliable Channel API
This API considers the types defined by MESSAGING-API plus the following.
Type definitions
types:
ReliableSendId:
type: string
description: "Unique identifier for a single `send` operation on a reliable channel.
It groups all chunks produced by segmenting one envelope, so callers can correlate
acknowledgement and error events back to the original send call.
Internally, each chunk is dispatched as an independent [MESSAGING-API](/standards/application/messaging-api.md) call,
producing one `RequestId` (as defined in [MESSAGING-API](/standards/application/messaging-api.md)) per chunk.
A single `ReliableSendId` therefore maps to one or more underlying `RequestId` values,
one per chunk sent."
IEncryption:
type: object
description: "Interface for a pluggable encryption mechanism.
When provided, see ReliableChannelConfig, the API consumer MUST implement both encrypt and decrypt operations.
Implementations MAY use different signatures than those described below, as long as each operation accepts a byte array and returns a byte array."
fields:
encrypt:
type: function
description: "Encrypts a byte payload. Returns the encrypted payload."
parameters:
- name: content
type: array<byte>
returns:
type: result<array<byte>, error>
decrypt:
type: function
description: "Decrypts a byte payload. Returns the decrypted payload."
parameters:
- name: payload
type: array<byte>
returns:
type: result<array<byte>, error>
IPersistence:
type: object
description: "Interface for a pluggable SDS persistence backend.
Implementations MUST provide all functions required to save and retrieve SDS state per channel. Implementations MUST also provide the persistence method of interest, e.g., SQLite, custom encrypted storage, etc.
Refer to the [SDS spec](https://lip.logos.co/ift-ts/raw/sds.html) for the full definition of what state must be persisted."
ReliableEnvelope:
type: object
description: "Wraps a MessageEnvelope (defined in [MESSAGING-API](/standards/application/messaging-api.md)) with a reliable channel identifier.
An empty channelId marks the message as ephemeral; a non-empty value routes it through the named reliable channel for segmentation, SDS tracking, and encryption."
fields:
messageEnvelope:
type: MessageEnvelope
description: "The message payload and metadata. Refer to [MESSAGING-API](/standards/application/messaging-api.md) for details."
channelId:
type: string
default: ""
description: "Reliable channel identifier.
If empty, the message is ephemeral: not tracked by SDS, never retransmitted, and NEVER segmented.
Ephemeral payloads exceeding the network size limit MUST be rejected with an error.
When the rate limit is approached, ephemeral messages are dropped immediately rather than queued.
If non-empty, the messageEnvelope is segmented, registered with SDS, and encrypted under the named reliable channel."
ReliableChannel:
type: object
description: "An entity that guarantees message delivery among all participants on a single content topic."
fields:
channelId:
type: string
description: "Unique identifier"
messageEvents:
type: MessageEvents
description: "Event emitter for message-related events on this channel"
ReliableChannelConfig:
type: object
fields:
segmentationConfig:
type: SegmentationConfig
description: "Configuration for message segmentation. Refer to [SEGMENTATION-API](./segmentation-api.md) for details."
sdsConfig:
type: SdsConfig
description: "Configuration for Scalable Data Sync."
rateLimitConfig:
type: RateLimitConfig
default: DefaultRateLimitConfig
description: "Configuration for rate limit management."
encryption:
type: optional<IEncryption>
default: none
description: "Optional pluggable encryption implementation. If none, messages are sent unencrypted."
SdsConfig:
type: object
fields:
persistence:
type: IPersistence
description: "Backend for persisting the SDS local history. Implementations MAY support custom backends."
acknowledgementTimeoutMs:
type: uint
default: 5000
description: "Time in milliseconds to wait for acknowledgement before retransmitting."
maxRetransmissions:
type: uint
default: 5
description: "Maximum number of retransmission attempts before considering delivery failed."
causalHistorySize:
type: uint
default: 2
description: "Number of message IDs to consider in the causal history. With longer value, a stronger correctness is guaranteed but it requires higher bandwidth and memory."
RateLimitConfig:
type: object
fields:
enabled:
type: bool
default: true
description: "Whether rate limiting is enforced. SHOULD be true when RLN is active."
epochSizeMs:
type: uint
default: 600000 # 10 minutes
description: "The epoch size used by the RLN relay, in milliseconds."
MessageReceivedEvent:
type: object
description: "Event emitted when a complete message has been received and reassembled."
fields:
requestId:
type: ReliableSendId
description: "Identifier of the `send` operation that produced this message."
envelope:
type: ReliableEnvelope
description: "The reassembled message and its channel context. `envelope.messageEnvelope.payload` contains the fully reassembled content; `envelope.channelId` identifies the channel on which it arrived."
MessageSentEvent:
type: object
description: "Event emitted when all chunks of a message have been acknowledged by the network."
fields:
requestId:
type: ReliableSendId
description: "The identifier of the `send` operation whose chunks have all been acknowledged."
MessageSendErrorEvent:
type: object
description: "Event emitted when a message send operation fails after exhausting retransmission attempts."
fields:
requestId:
type: ReliableSendId
description: "The identifier of the `send` operation that failed after exhausting retransmission attempts."
error:
type: string
description: "Error message describing what went wrong"
MessageEvents:
type: event_emitter
description: "Event source for reliable message events on a channel"
events:
"reliable:message:received":
type: MessageReceivedEvent
"reliable:message:sent":
type: MessageSentEvent
"reliable:message:send-error":
type: MessageSendErrorEvent
Channel lifecycle
functions:
createReliableChannel:
description: "Opens a reliable channel. Sets up the required SDS state, segmentation, and encryption."
parameters:
- name: channelConfig
type: ReliableChannelConfig
description: "Configuration for the channel."
- name: senderId
type: string
description: "An identifier for this sender. SHOULD be unique and persisted between sessions."
returns:
type: result<ReliableChannel, error>
closeChannel:
description: "Closes a reliable channel and releases all associated resources and internal state."
parameters:
- name: channel
type: ReliableChannel
description: "The channel to close."
returns:
type: result<void, error>
Channel usage
functions:
send:
description: "Send a message through the reliable channel API. Routing and guarantees are determined by the envelope's channelId: empty means ephemeral; non-empty applies segmentation, SDS, and encryption (if configured) before dispatching."
parameters:
- name: envelope
type: ReliableEnvelope
description: "The envelope containing the message and channel routing information."
returns:
type: result<ReliableSendId, error>
description: "Returns a `ReliableSendId` that callers can use to correlate subsequent `MessageSentEvent` or `MessageSendErrorEvent` events."
Incoming events are emitted on ReliableChannel.messageEvents as defined by MessageEvents.
Security/Privacy Considerations
- This API does not provide confidentiality by default. An
IEncryptionimplementation MUST be supplied when confidentiality is required. - Chunk metadata (message ID, chunk index, total chunks) is visible to network observers unless encrypted by the hook.
- SDS acknowledgement messages are sent over the same content topic and are subject to the same confidentiality concerns.
- Rate limiting compliance is required to avoid exclusion from the network by RLN-enforcing relays.
Copyright
Copyright and related rights waived via CC0.