Currently, the `connection_keep_alive` function of identify does not compute anything but its return value is set through the handler state machine. This is hard to understand and causes surprising behaviour because at the moment, we set `KeepAlive::No` as soon as the remote has answered our identify request. Depending on what else is happening on the connection, this might close the connection before we have successfully answered the remote's identify request.
To fix this, we now compute `connection_keep_alive` based on whether we are still using the connection.
Related: #3844.
Pull-Request: #3876.
The currently provided `ConnectionHandlerUpgrErr` is very hard to use. Not only does it have a long name, it also features 3 levels of nesting which results in a lot of boilerplate. Last but not least, it exposes `multistream-select` as a dependency to all protocols.
We fix all of the above by renaming the type to `StreamUpgradeError` and flattening out its interface. Unrecoverable errors during protocol selection are hidden within the `Io` variant.
Related: #3759.
Pull-Request: #3882.
This removes the deprecated `IntoConnectionHandler` trait and all its implementations. Consequently, `NetworkBehaviour::new_handler` and `NetworkBehaviour::addresses_of_peer` are now gone and the two `handle_` functions are now required to implement.
Related: #3647.
Pull-Request: #3884.
This patch refactors the identify tests to use `libp2p-swarm-test`. This allows us to delete quite a bit of code and makes several dev-dependencies obsolete.
The `correct_transfer` test is made obsolete by more precise assertions in the `periodic_identify` test. This allows us to remove the dependency on the `upgrade::{apply_inbound,apply_outbound}` functions.
Finally, we also fix a bug where the reported listen addresses to the other node could contain duplicates.
Related: #3748.
Pull-Request: #3851.
Previously, a protocol could be any sequence of bytes as long as it started with `/`. Now, we directly parse a protocol as `String` which enforces it to be valid UTF8.
To notify users of this change, we delete the `ProtocolName` trait. The new requirement is that users need to provide a type that implements `AsRef<str>`.
We also add a `StreamProtocol` newtype in `libp2p-swarm` which provides an easy way for users to ensure their protocol strings are compliant. The newtype enforces that protocol strings start with `/`. `StreamProtocol` also implements `AsRef<str>`, meaning users can directly use it in their upgrades.
`multistream-select` by itself only changes marginally with this patch. The only thing we enforce in the type-system is that protocols must implement `AsRef<str>`.
Resolves: #2831.
Pull-Request: #3746.
Previously, and for unknown legacy reasons, we waited for a configurable delay (default 500ms) upon new connections before we ran the identify protocol. This unnecessarily slows down applications that wait for the identify handshake to complete before performing further actions.
Resolves#3485.
Pull-Request: #3545.
Previously, we would implicitly establish a connection when the user wanted to push identify information to a peer. I believe that this is the wrong behaviour. Instead, I am suggesting to log a message that we are skipping the push to this peer.
Additionally, the way this is currently implemented does not make much sense. Dialing a peer takes time. In case we don't have a connection at all, it could be that we drop the push requests because there isn't an active handler and thus we would have unnecessarily established the connection.
Instead of fixing this - which would require buffering the push messages - I think we should just remove the implicit dial.
Pull-Request: #3843.
In the libp2p specs, the only handshake pattern that is specified is the XX handshake. Support for other handshake patterns can be added through external modules. While we are at it, we rename the remaining types to following the laid out naming convention.
The tests for handshakes other than XX are removed. The handshakes still work as we don't touch them in this patch.
Related #2217.
Pull-Request: #3768.
The `unreachable_pub` lint makes us aware of uses of `pub` that are not actually reachable from the crate root. This is considered good because it means reading a `pub` somewhere means it is actually public API. Some of our crates are quite large and keeping their entire API surface in your head is difficult.
We should strive for most items being `pub(crate)`. This lint helps us enforce that.
Pull-Request: #3735.
This PR renames some method names that don't follow Rust naming conventions or behave differently from what the name suggests:
- Enforce "try" prefix on all methods that return `Result`.
- Enforce "encode" method name for methods that return encoded bytes.
- Enforce "to_bytes" method name for methods that return raw bytes.
- Enforce "decode" method name for methods that convert encoded key.
- Enforce "from_bytes" method name for methods that convert raw bytes.
Pull-Request: #3775.
Mark constructors `Swarm::with_X_executor` as deprecated.
Move the deprecated functionality to `SwarmBuilder::with_X_executor`
Use `SwarmBuilder` throughout.
Resolves#3186.
Resolves#3107.
Pull-Request: #3588.
Instead of relying on `protoc` and buildscripts, we generate the bindings using `pb-rs` and version them within our codebase. This makes for a better IDE integration, a faster build and an easier use of `rust-libp2p` because we don't force the `protoc` dependency onto them.
Resolves#3024.
Pull-Request: #3312.
Previously, a `ConnectionHandler` was immediately requested from the `NetworkBehaviour` as soon as a new dial was initiated or a new incoming connection accepted.
With this patch, we delay the creation of the handler until the connection is actually established and fully upgraded, i.e authenticated and multiplexed.
As a consequence, `NetworkBehaviour::new_handler` is now deprecated in favor of a new set of callbacks:
- `NetworkBehaviour::handle_pending_inbound_connection`
- `NetworkBehaviour::handle_pending_outbound_connection`
- `NetworkBehaviour::handle_established_inbound_connection`
- `NetworkBehaviour::handle_established_outbound_connection`
All callbacks are fallible, allowing the `NetworkBehaviour` to abort the connection either immediately or after it is fully established. All callbacks also receive a `ConnectionId` parameter which uniquely identifies the connection. For example, in case a `NetworkBehaviour` issues a dial via `NetworkBehaviourAction::Dial`, it can unambiguously detect this dial in these lifecycle callbacks via the `ConnectionId`.
Finally, `NetworkBehaviour::handle_pending_outbound_connection` also replaces `NetworkBehaviour::addresses_of_peer` by allowing the behaviour to return more addresses to be used for the dial.
Resolves#2824.
Pull-Request: #3254.
Don't close the stream `protocol::recv`.
This is a short-term fix for #3298.
The issue behind this is a general one on the QUIC transport when closing streams, as described in #3343. This PR only circumvents the issue for identify. A proper solution for our QUIC transport still needs more thought.
Pull-Request: #3344.
We create the `ConnectionId` for the new connection as part of `DialOpts`. This allows `NetworkBehaviour`s to accurately track state regarding their own dial attempts.
This patch is the main enabler of https://github.com/libp2p/rust-libp2p/pull/3254. Removing the `handler` field will allow us to deprecate the `NetworkBehaviour::new_handler` function in favor of four new ones that give more control over the connection lifecycle.
Previously, we used the full reference to the `OutEvent` of the `ConnectionHandler` in all implementations of `NetworkBehaviour`. Not only is this very verbose, it is also more brittle to changes. With the current implementation plan for #2824, we will be removing the `IntoConnectionHandler` abstraction. Using a type-alias to refer to the `OutEvent` makes the migration much easier.
The trick with this one is to use `futures::Either` everywhere where we may wrap something that implements any of the `futures` traits. This includes the output of `EitherFuture` itself. We also need to implement `StreamMuxer` on `future::Either` because `StreamMuxer`s may be the the `Output` of `InboundUpgrade`.
With this commit `libp2p-identify` no longer discards the whole identify payload in case a listen addr of the remote node is invalid, but instead logs the failure, skips the invalid multiaddr and parses the remaining identify payload.
This is especially relevant when rolling out a new protocol to a live network. Say that most nodes of a network run on an implementation version v1. Say that the `multiaddr` implementation is not aware of the `webrtc/` protocol. Say that a new version (v2) is rolled out to the network with support for the `webrtc/` protocol, listening via `webrtc/` by default. In such case all v1 nodes would discard all identify payloads of v2 nodes, given that the v2 identify payloads would contain the `webrtc/` protocol in their `listen_addr` addresses.
See https://github.com/libp2p/rust-libp2p/issues/3244 for details.
This patch deprecates 3 out of 4 functions on `PollParameters`:
- `local_peer_id`
- `listened_addresses`
- `external_addresses`
The addresses can be obtained by inspecting the `FromSwarm` event. To make this easier, we introduce two utility structs in `libp2p-swarm`:
- `ExternalAddresses`
- `ListenAddresses`
A node's `PeerId` is always known to the caller, thus we can require them to pass it in.
Related: #3124.
Previously, we had one callback for each kind of message that a `ConnectionHandler` would receive from either its `NetworkBehaviour` or the connection itself.
With this patch, we combine these functions, resulting in two callbacks:
- `on_behaviour_event`
- `on_connection_event`
Resolves#3080.
Previously, the executor for connection tasks silently defaulted to a `futures::executor::ThreadPool`. This causes issues such as https://github.com/libp2p/rust-libp2p/issues/2230.
With this patch, we force the user to choose, which executor they want to run the connection tasks on which results in overall simpler API with less footguns.
Closes#3068.
* misc/metrics: Explicitly delegate event recording to each recorder
This allows delegating a single event to multiple `Recorder`s. That enables e.g. the
`identify::Metrics` `Recorder` to act both on `IdentifyEvent` and `SwarmEvent`. The latter enables
it to garbage collect per peer data on disconnects.
* protocols/dcutr: Expose PROTOCOL_NAME
* protocols/identify: Expose PROTOCOL_NAME and PUSH_PROTOCOL_NAME
* protocols/ping: Expose PROTOCOL_NAME
* protocols/relay: Expose HOP_PROTOCOL_NAME and STOP_PROTOCOL_NAME
* misc/metrics: Track # connected nodes supporting specific protocol
An example metric exposed with this patch:
```
libp2p_identify_protocols{protocol="/ipfs/ping/1.0.0"} 10
```
This implies that 10 of the currently connected nodes support the ping protocol.
Remove the concept of individual `Transport::Listener` streams from `Transport`.
Instead the `Transport` is polled directly via `Transport::poll`. The
`Transport` is now responsible for driving its listeners.
**Summary** of the plot of the `discover_peer_after_disconnect` test:
1. `swarm2` connects to `swarm1`.
2. `swarm2` requests an identify response from `swarm1`.
3. `swarm1` sends the response to `swarm2`.
4. `swarm2` disconnects from `swarm1`.
5. `swarm2` tries to disconnect.
**Problem**
`libp2p-identify` sets `KeepAlive::No` when it identified the remote. Thus `swarm1` might
identify` `swarm2` before `swarm2` identified `swarm1`. `swarm1` then sets `KeepAlive::No` and thus closes the
connection to `swarm2` before `swarm2` identified `swarm1`. In such case the unit test
`discover_peer_after_disconnect hangs indefinitely.
**Solution**
Add an initial delay to `swarm1` requesting an identification from `swarm2`, thus ensuring `swarm2`
is always able to identify `swarm1`.
An identify push contains the whole identify information of a remote
peer. Upgrading multiple inbound identify push streams is useless.
Instead older streams are dropped in favor of newer streams.
