This is now a very simple option serving multiple purposes:
* It allows for stable (integration) tests involving a Swarm, which
are otherwise subject to race conditions due to the connection being
allowed to terminate at any time with `KeepAlive::No`
(which remains the default).
* It makes for a more entertaining ping example which continuously
sends pings.
* Maybe someone wants to use the ping protocol for application-layer
connection keep-alive after all.
* Fix connection & handler shutdown when using `KeepAlive::Now`.
Delay::new(Instant::now()) is never immediately ready, resulting in
`KeepAlive::Now` to have no effect, since the delay is re-created on
every execution of `poll()` in the `NodeHandlerWrapper`. It can also
send the node handler into a busy-loop, since every newly
created Delay will trigger a task wakeup, which creates a new Delay
with Instant::now(), and so forth.
The use of `Delay::new(Instant::now())` for "immediate" connection shutdown
is therefore removed here entirely. An important assumption is thereby
that as long as the node handler non-empty `negotiating_in` and `negotiating_out`,
the handler is not dependent on such a Delay for task wakeup.
* Correction to the libp2p-ping connection timeout.
The current connection timeout is always short of one `interval`,
because the "countdown" begins with the last received or sent pong
(depending on the policy). In effect, the current default config has
a connection timeout of 5 seconds (20 - 15) from the point when a ping is sent.
Instead, the "countdown" of the connection timeout should always begin
with the next scheduled ping. That also makes all configurations valid,
avoiding pitfalls.
The important properties of the ping handler are now checked to hold for all
configurations, in particular:
* The next ping must be scheduled no earlier than the ping interval
and no later than the connection timeout.
* The "countdown" for the connection timeout starts on the next ping,
i.e. the full connection timeout remains at the instant when the
next ping is sent.
* Do not keep connections alive.
The ping protocol is not supposed to keep otherwise idle connections
alive, only to add an additional condition for terminating them in
the form of a configurable number of consecutive failed ping requests.
In this context, the `PingPolicy` does not seem useful any longer.
* libp2p-ping improvements.
* re #950: Removes use of the `OneShotHandler`, but still sending each
ping over a new substream, as seems to be intentional since #828.
* re #842: Adds an integration test that exercises the ping behaviour through
a Swarm, requiring the RTT to be below a threshold. This requires disabling
Nagle's algorithm as it can interact badly with delayed ACKs (and has been
observed to do so in the context of the new ping example and integration test).
* re #864: Control of the inbound and outbound (sub)stream protocol upgrade
timeouts has been moved from the `NodeHandlerWrapperBuilder` to the
`ProtocolsHandler`. That may also alleviate the need for a custom timeout
on an `OutboundSubstreamRequest` as a `ProtocolsHandler` is now free to
adjust these timeouts over time.
Other changes:
* A new ping example.
* Documentation improvements.
* More documentation improvements.
* Add PingPolicy and ensure no event is dropped.
* Remove inbound_timeout/outbound_timeout.
As per review comment, the inbound timeout is now configured
as part of the `listen_protocol` and the outbound timeout as
part of the `OutboundSubstreamRequest`.
* Simplify and generalise.
Generalise `ListenProtocol` to `SubstreamProtocol`, reusing it in
the context of `ProtocolsHandlerEvent::OutboundSubstreamRequest`.
* Doc comments for SubstreamProtocol.
* Adapt to changes in master.
* Relax upper bound for ping integration test rtt.
For "slow" CI build machines?
