With https://github.com/libp2p/rust-libp2p/pull/2248 a connection task
`await`s sending an event to the behaviour before polling for new events
from the behaviour [1].
When `Swarm::poll` is unable to deliver an event to a connection task it
returns `Poll::Pending` even though (a) polling `Swarm::network` might be
able to make progress (`network_not_ready` being `false`) and (b) it
does not register a waker to be woken up [2].
In combination this can lead to a deadlock where a connection task waits
to send an event to the behaviour and `Swarm::poll` returns
`Poll::Pending` failing to send an event to the connection task, not
registering a waiker in order to be polled again.
With this commit `Swarm::poll` will only return `Poll::Pending`, when
failing to deliver an event to a connection task, if the network is
unable to make progress (i.e. `network_not_ready` being `true`).
In the long-run `Swarm::poll` should likely be redesigned, prioritizing
the behaviour over the network, given the former is the control plane
and the latter potentially yields new work from the outside.
[1]: ca1b7cf043/core/src/connection/pool/task.rs (L224-L232)
[2]: ca1b7cf043/swarm/src/lib.rs (L756-L783)
The test `max_established_incoming` starts two networks with a
configured connection limit, spawns up to `limit + 1` connections and
expects the last connection to be closed due to being over the limit.
The previous test implementation depended on both networks to handle
each connection in sequence, which is not always the case.
E.g. while network 2 might expect the last connection to close, network
1 might finish upgrading the last connection before the second to last
connection, thus expecting the second to last connection to close.
This commit drives network 1 and 2 in sequence and ensures both networks
are finished upgrading a connection before starting a new connection. In
addition it upgrade the test to use `quickcheck`.
Concurrently dial address candidates within a single dial attempt.
Main motivation for this feature is to increase success rate on hole punching
(see https://github.com/libp2p/rust-libp2p/issues/1896#issuecomment-885894496
for details). Though, as a nice side effect, as one would expect, it does
improve connection establishment time.
Cleanups and fixes done along the way:
- Merge `pool.rs` and `manager.rs`.
- Instead of manually implementing state machines in `task.rs` use
`async/await`.
- Fix bug where `NetworkBehaviour::inject_connection_closed` is called without a
previous `NetworkBehaviour::inject_connection_established` (see
https://github.com/libp2p/rust-libp2p/issues/2242).
- Return handler to behaviour on incoming connection limit error. Missed in
https://github.com/libp2p/rust-libp2p/issues/2242.
Create a `ListenersEvent::Closed` when a listener is removed via
`Swarm::remove_listener`. This makes it more consistent with `Swarm::listen_on`,
and also informs the Swarm about the associated expired addresses.
Co-authored-by: Max Inden <mail@max-inden.de>
Set "openmetrics-text" content type on HTTP GET response. Makes sure
Prometheus server parses returned metrics via OpenMetrics format instead
of legacy Prometheus format.
To have caches operate at its maximum usefulness, we need to have a
1-to-1 mapping between build-comamnd and cache key. Otherwise rustc
has to rebuild certain artifacts because they were not in the cache.
By using matrices, we make github parallelise some of our jobs. This
has a double positive impact on CI runtime. Not only are our caches
more effective now, several jobs are now run in parallel.
Co-authored-by: Max Inden <mail@max-inden.de>
Previously, peers that did not support gossipsub were removed from the
connection-id mappings.
This can cause the connection_id mappings to go out of sync with those managed
by the swarm. This PR corrects this and adds an extra event that can inform the
user that a peer that does not support the protocol has connected. The user can
then optionally handle peers that don't support the protocol.
* Add support for the `Identify` protocol to the server, such that the
`register_with_identify` example works as intended
* Add discovery loop to the `discovery` example and demonstrate cookie
usage
* Drop explicit dependency on async_std
Co-authored-by: Max Inden <mail@max-inden.de>
Probability for a Kademlia query to return in less than 100 milliseconds
is low, thus increasing the lower bucket to improve accuracy within the
higher ranges.
Implement the libp2p rendezvous protocol.
> A lightweight mechanism for generalized peer discovery. It can be used for
bootstrap purposes, real time peer discovery, application specific routing, and
so on.
Co-authored-by: rishflab <rishflab@hotmail.com>
Co-authored-by: Daniel Karzel <daniel@comit.network>
Require `NetworkBehaviourAction::{DialPeer,DialAddress}` to contain a
`ProtocolsHandler`. This allows a behaviour to attach custom state to its
handler. The behaviour would no longer need to track this state separately
during connection establishment, thus reducing state required in a behaviour.
E.g. in the case of `libp2p-kad` the behaviour can include a `GetRecord` request
in its handler, or e.g. in the case of `libp2p-request-response` the behaviour
can include the first request in the handler.
Return `ProtocolsHandler` on connection error and close. This allows a behaviour
to extract its custom state previously included in the handler on connection
failure and connection closing. E.g. in the case of `libp2p-kad` the behaviour
could extract the attached `GetRecord` from the handler of the failed connection
and then start another connection attempt with a new handler with the same
`GetRecord` or bubble up an error to the user.
Co-authored-by: Thomas Eizinger <thomas@eizinger.io>
With f2905c07f1246c3c3fdc1cde95f7e9c5c1c9b01a the secp256k1 feature is
disabled by default. Instead of enabling it in the dev-dependency,
simply use ed25519.
