* refactor globals snapshot
* ignore test
* update pwasm-utils ref
* line width
* add doc comment for internal struct
* add explanation for iteration
* Demote rustdoc to a comment
* use 0.14
Co-authored-by: Sergei Shulepov <sergei@parity.io>
* Rough skeleton for what I think the RPC should look like
* Create channel for sending justifications
Sends finalized header and justification from Grandpa to the
client. This lays the groundwork for hooking into the RPC module.
* WIP: Add subscribers for justifications to Grandpa
Adds the Sender end of a channel into Grandpa, through which notifications
about block finality events can be sent.
* WIP: Add a struct for managing subscriptions
Slightly different approach from the last commit, but same
basic idea. Still a rough sketch, very much doesn't compile yet.
* Make naming more clear and lock data in Arc
* Rough idea of what RPC would look like
* Remove code from previous approach
* Missed some things
* Update client/rpc-api/src/chain/mod.rs
Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com>
* Update client/rpc-api/src/chain/mod.rs
Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com>
* Split justification subscription into sender and receiver halves
* Replace RwLock with a Mutex
* Add sample usage from the Service's point of view
* Remove code that referred to "chain_" RPC
* Use the Justification sender/receivers from Grandpa LinkHalf
* Add some PubSub boilerplate
* Add guiding comments
* TMP: comment out to fix compilation
* Return MetaIoHandler from PubSubHandler in create_full
* Uncomment pubsub methods in rpc handler (fails to build)
* node/rpc: make Metadata concrete in create_full to fix compilation
* node: pass in SubscriptionManger to grandpa rpc handler
* grandpa-rpc: use SubscriptionManger to add subscriber
* grandpa-rpc: attempt at setting up the justification stream (fails to build)
* grandpa-rpc: fix compilation of connecting stream to sink
* grandpa-rpc: implement unsubscribe
* grandpa-rpc: update older tests
* grandpa-rpc: add full prefix to avoid confusing rust-analyzer
* grandpa-rpc: add test for pubsub not available
* grandpa-rpc: tidy up leftover code
* grandpa-rpc: add test for sub and unsub of justifications
* grandpa-rpc: minor stylistic changes
* grandpa-rpc: split unit test
* grandpa-rpc: minor stylistic changes in test
* grandpa-rpc: skip returning future when cancelling
* grandpa-rpc: reuse testing executor from sc-rpc
* grandpa-rpc: don't need to use PubSubHandler in tests
* node-rpc: use MetaIoHandler rather than PubSubHandler
* grandpa: log if getting header failed
* grandpa: move justification channel creation into factory function
* grandpa: make the justification sender optional
* grandpa: fix compilation warnings
* grandpa: move justification notification types to new file
* grandpa-rpc: move JustificationNotification to grandpa-rpc
* grandpa-rpc: move JustificationNotification to its own file
* grandpa: rename justification channel pairs
* grandpa: rename notifier types
* grandpa: pass justification as GrandpaJustification to the rpc module
* Move Metadata to sc-rpc-api
* grandpa-rpc: remove unsed error code
* grandpa: fix bug for checking if channel is closed before sendind
* grandpa-rpc: unit test for sending justifications
* grandpa-rpc: update comments for the pubsub test
* grandpa-rpc: update pubsub tests with more steps
* grandpa-rpc: fix pubsub test
* grandpa-rpc: minor indendation
* grandpa-rpc: decode instead of encode in test
* grandpa: fix review comments
* grandpa: remove unused serde dependency
Co-authored-by: Tomasz Drwięga <tomusdrw@users.noreply.github.com>
Co-authored-by: Jon Häggblad <jon.haggblad@gmail.com>
Co-authored-by: Tomasz Drwięga <tomasz@parity.io>
* Revalidate transactions only on latest best block
We should revalidate transactions only on the latest best block and not
on any arbitrary block. The revalidation before failed when there were
multiple blocks on the height given to the revalidation function, but no
block was imported as best block.
* Update test-utils/runtime/transaction-pool/src/lib.rs
Co-authored-by: Jaco Greeff <jacogr@gmail.com>
* Fix tests
* Only process best blocks in the transaction pool
Co-authored-by: Jaco Greeff <jacogr@gmail.com>
* Allow blacklisting blocks from being finalized again after block revert
* Use BlockRules for storing unfinalized and add have_state_at in revert
* Move finalization_check in finalize_block upward
* Directly mark finalization blacklist as badblocks
* Remove obselete comment
* BabeWorker -> BabeSlotWorker
* SlotWorker::notify_slot: similar to claim_slot, but called no matter authoring
* Wrap the future with a new struct BabeWorker
* Add type definition slot_notification_sinks
* Function slot_notification_streams for the receiver side
* Get a handle of slot_notification_sinks in BabeSlotWorker
* Implement notify_slot
* Switch to use bounded mpsc
* Do not drop the sink when channel is full
Only skip sending the message and emit a warning, because it is recoverable.
* Fix future type bounds
* Add must_use and sink type alias
* client/network/src/discovery: Adjust to Kademlia API changes
* client/network: Add peers to DHT only if protocols match
With https://github.com/libp2p/rust-libp2p/pull/1628 rust-libp2p allows
manually controlling which peers are inserted into the routing table.
Instead of adding each peer to the routing table automatically, insert
them only if they support the local nodes protocol id (e.g. `dot`)
retrieved via the `identify` behaviour.
For now this works around
https://github.com/libp2p/rust-libp2p/issues/1611. In the future one
might add more requirements. For example one might try to exclude
light-clients.
* Cargo.toml: Remove crates.io patch for libp2p
* client/network/src/behaviour: Adjust to PeerInfo name change
* client/network/src/discovery: Rework Kademlia event matching
* client/network/discovery: Add trace on adding peer to DHT
* client/network/discovery: Retrieve protocol name from kad behaviour
* client/network/discovery: Fix formatting
* client/network: Change DiscoveryBehaviour::add_self_reported signature
* client/network: Document manual insertion strategy
* client/network/discovery: Remove TODO for ignoring DHT address
Co-authored-by: Pierre Krieger <pierre.krieger1708@gmail.com>
* Add NetworkService::send_notifications
* Doc
* Doc
* API adjustment
* Address concerns
* Make it compile
* Start implementation
* Progress in the implementation
* Change implementation strategy again
* More work before weekend
* Finish changes
* Minor doc fix
* Revert some minor changes
* Apply suggestions from code review
* GroupError -> NotifsHandlerError
* Apply suggestions from code review
Co-authored-by: Roman Borschel <romanb@users.noreply.github.com>
* state_transition_waker -> close_waker
* Apply suggestions from code review
Co-authored-by: Roman Borschel <romanb@users.noreply.github.com>
* Finish renames in service.rs
* More renames
* More review suggestsions applied
* More review addressing
* Final change
* 512 -> 2048
Co-authored-by: Roman Borschel <romanb@users.noreply.github.com>
* Remove service components and add build_network, build_offchain_workers etc
* Improve transaction pool api
* Remove commented out line
* Add PartialComponents
* Add BuildNetworkParams, documentation
* Remove unused imports in tests
* Apply suggestions from code review
Co-authored-by: Nikolay Volf <nikvolf@gmail.com>
* Remove unused imports in node-bench
Co-authored-by: Nikolay Volf <nikvolf@gmail.com>
* grandpa: remove unused methods to convert digest
* grandpa: add root extrinsic for scheduling forced change
* grandpa: add benchmark for schedule_forced_change
* grandpa: don't take authority weight in schedule_forced_change
* grandpa: add const for default forced change delay
* grandpa: adjust weights after benchmark on ref hardware
* grandpa: fix cleanup of forced changes on standard change application
* grandpa: replace schedule_forced_change with note_stalled
* grandpa: always trigger a session change when the set is stalled
* grandpa: fix bug on set id mutation after failed scheduled change
* grandpa: take delay as parameter in note_stalled
* grandpa: fix tests
* grandpa: fix cleanup of forced changes
* grandpa: add test for forced changes cleanup
* grandpa: add test for session rotation set id
* grandpa: add test for scheduling of forced changes on new session
* Ensure we close old file handles hold by sysinfo
* Dropping is needed unfortunately
* enable process refreshing, ignore result from refresh_process
* jumping to proposed patch
* switch to latest sysinfo
* Switch to serde_json::Value for Visitor values
* add tests
* switch Visitor to use Map
* refactor change names
* \n
* update integration test
* use discrete maps for each type of recorded value
* update integration test
* add docs
* Update client/tracing/src/lib.rs
Co-authored-by: Kian Paimani <5588131+kianenigma@users.noreply.github.com>
* add docs
Co-authored-by: Matt Rutherford <mattrutherford@users.noreply.github.com>
Co-authored-by: Kian Paimani <5588131+kianenigma@users.noreply.github.com>
* Initial commit
Forked at: 74655edebc
Parent branch: origin/master
* Move task_manager.rs to mod.rs
* Graceful shutdown for the task manager
* Await all background task JoinHandle at the same time
* Add tests
* Make future() wait also for exit signal + fix essential task failed
Probably related to https://github.com/paritytech/cumulus/issues/111
* add comments for non-obvious code
* Use clean_shutdown() in sc-cli
* Adapt code and upgrade tokio in sc-cli
* cleanup spacing in doc
* Add license
* I guess actually running the clean shutdown would be a good idea
* fix tests
* Update client/cli/src/runner.rs
Co-authored-by: Benjamin Kampmann <ben@gnunicorn.org>
* Improve error logging
* disable other tests (can't reproduce on my machine)
* Revert "disable other tests (can't reproduce on my machine)"
This reverts commit c133c590f33c253123ba0555ce719a71ededd60d.
* It is possible that the tasks are ended first
* Revert "It is possible that the tasks are ended first"
This reverts commit 502aba4a49fb5d892e704c412b8a81768a3f2c71.
* Use single threaded scheduler for more predictability
* enable_time
* Revert "enable_time"
This reverts commit 4e152140764a4bddeedff06a4e36ec701909e8c7.
* Revert "Use single threaded scheduler for more predictability"
This reverts commit ee5e13c5f13ff71e012dcda13579dffeb15f8ffc.
* Revert "Revert "It is possible that the tasks are ended first""
This reverts commit 1b91a8ca3eebbdc18be199c8ca188e88669ae649.
* This cannot be verified either with a threaded pool
* Apply suggestions from code review
Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>
Co-authored-by: Benjamin Kampmann <ben@parity.io>
Co-authored-by: Benjamin Kampmann <ben@gnunicorn.org>
Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>
André Silva