This PR refactors the metrics measuring and Prometheus exposing entity in sc-service into its own submodule and extends the parameters it exposes by:
- system load average (over one, five and 15min)
- the TCP connection state of the process (lsof), refs #5304
- number of tokio threads
- number of known forks
- counter for items in each unbounded queue (with internal unbounded channels)
- number of file descriptors opened by this process (*nix only at this point)
- number of system threads (*nix only at this point)
refs #4679
Co-authored-by: Max Inden <mail@max-inden.de>
Co-authored-by: Ashley <ashley.ruglys@gmail.com>
* Split the Roles bitfield in three
* Forgot to include some changes
* Fix cli test
* More test fixes
* Oh God, merging master broke other tests
* Didn't run the doctests
* Address review
* I'm trying to fix the build blindly because it's taking a good hour to compile on my machine
* Address some review
* Also update the peerset's API to make sense
* Fix peerset tests
* Fix browser node
* client: distinguish between local and network authority
Co-authored-by: André Silva <andre.beat@gmail.com>
* client/network-gossip/src/bridge: Finish when network even stream closes
Previously within `<GossipEngine as Future>::poll` one would poll the
`network_event_stream` ignoring all messages other than
`Poll::Ready(Some())`. Ignoring `Poll::Ready(None)` leads to a panic on
the next poll of the stream, gien that it is not fused.
By design `network_event_stream` does not close unless an unbounded send
into it fails, or the `NetworkWorker` gets shut down.
> The stream never ends (unless the `NetworkWorker` gets shut down).
> (client/network/src/service.rs)
An `unbounded_send` to fail on an unbounded channel is unlikely. The
`NetworkWorker` shutting down is not unlikely. In such case the
`GossipEngine` should shut down as well.
With this patch a `<GossipEngine as Future>` finishes on
`Poll::Ready(None)` returned from `network_event_stream`.
* client/finality-grandpa/communication: Error on gossip engine finished
Have `<NetworkBridge as Future>::poll` return `Poll::Ready(Err)` instead
of `Poll::Ready(Ok)` to be consistent with the handling of the neighbor
packet worker stream and the gossip validator report stream. Both `Err`
as well as `Ok` shut down the `NetworkBridge` as well as the
`VoterWorker`.
* client/network-gossip/src/bridge: Add regression test
* client/network-gossip: Move substrate test client to dev dependencies
* client/network-gossip: Remove TODO
Addressed in a follow up pull request.
* client/network-gossip/bridge: Put match on newline after loop
* client/finality-grandpa/src/observer: Fix regression test
Make sure the event stream sender side is not dropped till the end.
* client/finality-grandpa: Add Prometheus metrics to GossipValidator
Instrument finality grandpa `GossipValidator` exposing count of messages
validated by message type and message action.
```
\# HELP substrate_finality_grandpa_communication_gossip_validator_messages Number of messages validated by the finality grandpa gossip validator.
\# TYPE substrate_finality_grandpa_communication_gossip_validator_messages counter
substrate_finality_grandpa_communication_gossip_validator_messages{action="discard",message="neighbor"} 39
substrate_finality_grandpa_communication_gossip_validator_messages{action="keep",message="vote"} 28
```
* client/finality-grandpa: Add None as Prometheus registry in tests
* client/finality-granpda/src/communication: Refactor metric registration
* Use CLI to configure max instances cache
* Fix tests
* Move default value into CLI
* Use SmallVec
* Apply review comments
* Get rid of `SmallVec`
Co-authored-by: Bastian Köcher <git@kchr.de>
* grandpa: only gossip commits to peers on the same set
* grandpa: track commits uniquely by round and set
* grandpa: fix communication test
* grandpa: add tests for commit gossip handling
* grandpa: add missing docs
* client/network-gossip: Merge GossipEngine and GossipEngineInner
Given that GossipEngine and GossipEngineInner are not shared between
threads anyone (public interface + background tasks), neither depends on
being Send or Sync. Thus one can merge the two as done in this patch.
One only needs to wrap an `Arc<Mutex<>>` around the whole structure when
the owner (e.g. finality-grandpa) needs to share the gossip engine
between threads.
* client/finality-grandpa: Wrap GossipEngine in Arc Mutex & lock it on use
GossipEngine in itself has no need to be Send and Sync, given that it
does not rely on separately spawned background tasks anymore. Given that
finality-grandpa shares the `NetworkBridge` potentially between threads
its components need to be clonable, thus this patch wraps `GossipEngine`
in an `Arc<Mutex<>>`.
* Open one substream for each notifications protocol
* Fix WASM build
* Apply suggestions from code review
Co-Authored-By: Toralf Wittner <tw@dtex.org>
* Address concerns
* Use unsigned-varint to read the varint
* Use unsigned-varint
* Forgot Cargo.lock
Co-authored-by: Toralf Wittner <tw@dtex.org>
`GossipEngine` spawns two tasks, one for a periodic tick, one to forward
messages from the network to subscribers. These tasks hold an `Arc` to a
`GossipEngineInner`.
To reduce the amount of shared ownership (locking) this patch integrates
the two tasks into a `Future` implementation on the `GossipEngine`
struct. This `Future` implementation can now be called from a single
owner, e.g. the `finality-grandpa` `NetworkBridge`.
As a side effect this removes the requirement on the `network-gossip`
crate to spawn tasks and thereby removes the requirement on the
`finality-grandpa` crate to spawn any tasks.
This is part of a greater effort to reduce the number of owners of
components within `finality-grandpa`, `network` and `network-gossip` as
well as to reduce the amount of unbounded channels. For details see
d9837d7dd, 5f80929dc and 597c0a6c4.
* Make WASM browser thing compile
* Fix
* updated exit-future (github repo)
* Switch to broadcast crate
* Migrate client/cli
* Switch exit-future to modernize branch
* Small changes
* Switch to cargo version and fix fg tests
* fix basic-authorship
* Fix crash on grafana macro
* Fix grafana macro
* Switch node python version
* Disable record_metrics_slice in grafana macro on wasm
* Update client/grafana-data-source/src/lib.rs
* Revert "Update client/grafana-data-source/src/lib.rs"
This reverts commit 888009a8e0b7051bd4bfbbfdb0448bcf2e2aae93.
* Add wasm support for state machine
* Switch to my own libp2p version
* Revert "Switch to my own libp2p version"
This reverts commit ce613871b59264b3165b45c37943e6560240daa7.
* Revert "Add wasm support for state machine"
This reverts commit de7eaa0694d9534fc3b164621737968e9a6a7c5f.
* Add sc-browser
* Squash
* remove sc-browser
* Fix keystore on wasm
* stubs for removed functions to make env compatible with old runtimes
* Add test (that doesn't work)
* Fix build scripts
* Revert basic-authorship due to no panics
* Revert cli/informant
* Revert consensus
* revert offchain
* Update utils/browser/Cargo.toml
Co-Authored-By: Benjamin Kampmann <ben@gnunicorn.org>
* export console functions
* Add new chainspec
* Fix ws in chain spec
* revert chainspec
* Fix chainspec
* Use an Option<PathBuf> in keystore instead of cfg flags
* Remove crud
* Only use wasm-timer for instant and systemtime
* Remove telemetry changes
* Assuming this is ok
* Add a KeystoreConfig
* Add stubs back in
* Update libp2p
* Revert "Add stubs back in"
This reverts commit 4690cf1882aa0f99f7f00a58c4080c8aa9b77c36.
* Remove commented js again
* Bump kvdb-web version
* Fix cli
* Switch branch on futures-timer
* Fix tests
* Remove sc-client test build in check-web-wasm because there isn't a good way to build futures-timer with wasm-bindgen support in the build
* Remove more things ^^
* Switch branch on futures-timer back
* Put DB io stats behind a cfg flag
* Fix things
* Don't timeout transports on wasm
* Update branch of futures-timer and fix bad merge
* Spawn informant
* Fix network test
* Fix delay resets
* Changes
* Fix tests
* use wasm_timer for transaction pool
* Fixes
* Switch futures-timer to crates
* Only diagnose futures on native
* Fix sc-network-test tests
* Select log level in js
* Fix syncing ;^)
* Allow disabling colours in the informant
* Use OutputFormat enum for informant
* MallocSizeOf impl on transaction pool broke stuff because wasm_timer::Instant doesnt impl it so just revert the transaction pool to master
* Update futures-diagnose
* Revert "MallocSizeOf impl on transaction pool broke stuff because wasm_timer::Instant doesnt impl it so just revert the transaction pool to master"
This reverts commit baa4ffc94fd968b6660a2c17ba8113e06af15548.
* Pass whole chain spec in start_client
* Get Instant::now to work in transaction pool again
* Informant dep reordering
Co-authored-by: Pierre Krieger <pierre.krieger1708@gmail.com>
Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>
Co-authored-by: Svyatoslav Nikolsky <svyatonik@gmail.com>
Co-authored-by: Benjamin Kampmann <ben.kampmann@googlemail.com>
Co-authored-by: Demi Obenour <48690212+DemiMarie-parity@users.noreply.github.com>
* client/finality-grandpa: Add regression test observer polling network
Ensure `Future` implementation of `ObserverWork` is polling its
`NetworkBridge`. Regression test for bug introduced in d9837d7dd and
fixed in 504b4e89e.
When polled, `NetworkBridge` forwards reputation change requests from
the `GossipValidator` to the underlying `dyn Network`. This test
triggers a reputation change by calling `GossipValidator::validate` with
an invalid gossip message. After polling the `ObserverWork` which should
poll the `NetworkBridge`, the reputation change should be forwarded to
the test network.
* Nits
Co-authored-by: Max Inden <mail@max-inden.de>
* Switch GrandPa to new futures
* Work on making tests work
* until_imported tests working again
* Work on switching tests to stable futures
* Modifications
* Re-add test as #[ignore]
* Don't ignore
* Add manual unpins
* Remove Header import
* Return concrete Sink type
* Switch to crates.io finality-grandpa version
* Remove use statement that slipped in
* Fix some nitpicks
* Remove unpin from i
* Fixed typo
* Move futures01 to dev-deps
* Fix nitpicks
* Update client/finality-grandpa/src/communication/mod.rs
Co-Authored-By: André Silva <andre.beat@gmail.com>
* nitpicking
Co-authored-by: Pierre Krieger <pierre.krieger1708@gmail.com>
Co-authored-by: André Silva <andre.beat@gmail.com>
* clinet/finality-grandpa: Make round_communication use bounded channel
`round_communication` returns a `Sink` and a `Stream` for outgoing and
incoming messages. The messages send into the `Sink` are forwarded down
to the network as well as send back into the `Stream` to ensure the node
processes its own messages.
So far, to send messages into the `Sink` back into the `Stream`, an
unbounded channel was used. This patch updates `round_communication` and
`OutgoingMessages` to use a bounded channel.
This is part of a greater effort to reduce the number of owners of
components within `finality-grandpa` and `network` as well as to reduce
the amount of unbounded channels. For details see d9837d7dd and
5f80929dc.
* client/finality-grandpa: Import futures03::future::ready at the top
* client/finality-grandpa: Make tests use compat of future 03
* client/finality-grandpa: Do not import ready into scope
Instead of importing futures03::future::ready into the scope, only
import futures::future03 into scope and call ready as furure03::ready.
* client/finality-grandpa: Reintegrate gossip validator report stream
The `finality-grandpa` `GossipValidator` is called by the `GossipEngine`
in a synchronous fashion on each gossip message. Its main task is to
decide whether to gossip the given message on, or whether to drop it.
In addition it also updates the reputation of a node's peers based on
the incoming gossip messages. To do so it needs to be able to report the
reputation change which it does through an unbounded channel (in order
to stay synchronous).
Previously the receiving side of this channel would be handled by a new
task, polling the channel and forwarding the changes to a clone of the
`GossipEngine` that it would own.
Instead the receiver of the above mentioned channel is now being polled
by the `NetworkBridge` within its `Future::poll` implementation.
Reputation changes are reported through the already existing
`GossipEngine` instance within `NetworkBridge`.
For details on the overall goal, see d9837d7dd.
* client/finality-grandpa: Remove exit future from test NetworkBridges
The `NeighborPacketWorker` within `client/finality-grandpa` does two
things:
1. It receives neighbor packets from components within
`client/finality-grandpa`, sends them down to the `GossipEngine` in
order for neighboring nodes to receive.
2. It periodically sends out the most recent neighbor packet to the
`GossipEngine`.
In order to send out packets it had a clone to a `GossipEgine` within
an atomic reference counter and a mutex. The `NeighborPacketWorker` was
then spawned onto its own asynchronous task.
Instead of running in its own task, this patch reintegrates the
`NeighborPacketWorker` into the main `client/finality-grandpa` task not
requiring the `NeighborPacketWorker` to own a clone of the
`GossipEngine`.
The greater picture
This is a tiny change within a greater refactoring. The overall goal is
to **simplify** how finality-grandpa interacts with the network and to
**reduce** the amount of **unbounded channels** within the logic.
Why no unbounded channels: Bounding channels is needed for backpressure
and proper scheduling. With unbounded channels there is no way of
telling the producer side to slow down for the consumer side to catch
up. Rephrased, there is no way for the scheduler to know when to favour
the consumer task over the producer task on a crowded channel and the
other way round for an empty channel.
Reducing the amount of shared ownership simplifies the logic and enables
one to use async-await syntax-suggar, given that one does not need to
hold a lock across poll invocations. Using async-await enables one to
use bounded channels without complex logic.
The finality-grandpa module needs two sets of functionalities from the
network:
1. Everything gossip related, e.g. event_stream, write_notification, ...
2. The ability to set a fork sync request for a specific block hash.
Instead of embedding (2) inside of (1) this patch extracts (2) from (1)
having finality-grandpa depend on a `Network` that fulfills the
`network_gossip::Network` trait and that can set block sync requests.
On the one hand this improves the overall structure splitting things
that don't logically belong together. On the other hand it does
reintroduce a lot of trait bounds within finality-grandpa.
* take test-client off sp-consensus
* use test primitives rather than test client in authority discovery tests
* move runtime-interface tests
* don't forget to remove the dev-dependency
* remove more unneeded dev deps
* add changes_trie_config to test prrimitives
* Separate network crates from its integration tests
* Fix up consensus crates for networking test changes
* remove unnecessary dependencies
* remove unused addition
* remove unnecessary dev-dependencies
* fixing finality grandpa tests
* removing unnecessary executor dependencies
* Adding script for rename, could be applicable for nodes on top of it, too
* add stderr and gitlab ci features
* apply script
* fix now minor details in expected stderr
* Update the Cargo.lock
* fix name: sc-transaction -> sc-tracing
* fix rename in script, too
* Adding first rough ouline of the repository structure
* Remove old CI stuff
* add title
* formatting fixes
* move node-exits job's script to scripts dir
* Move docs into subdir
* move to bin
* move maintainence scripts, configs and helpers into its own dir
* add .local to ignore
* move core->client
* start up 'test' area
* move test client
* move test runtime
* make test move compile
* Add dependencies rule enforcement.
* Fix indexing.
* Update docs to reflect latest changes
* Moving /srml->/paint
* update docs
* move client/sr-* -> primitives/
* clean old readme
* remove old broken code in rhd
* update lock
* Step 1.
* starting to untangle client
* Fix after merge.
* start splitting out client interfaces
* move children and blockchain interfaces
* Move trie and state-machine to primitives.
* Fix WASM builds.
* fixing broken imports
* more interface moves
* move backend and light to interfaces
* move CallExecutor
* move cli off client
* moving around more interfaces
* re-add consensus crates into the mix
* fix subkey path
* relieve client from executor
* starting to pull out client from grandpa
* move is_decendent_of out of client
* grandpa still depends on client directly
* lemme tests pass
* rename srml->paint
* Make it compile.
* rename interfaces->client-api
* Move keyring to primitives.
* fixup libp2p dep
* fix broken use
* allow dependency enforcement to fail
* move fork-tree
* Moving wasm-builder
* make env
* move build-script-utils
* fixup broken crate depdencies and names
* fix imports for authority discovery
* fix typo
* update cargo.lock
* fixing imports
* Fix paths and add missing crates
* re-add missing crates
André Silva