1. Refactored service manager with cleaner interface.
2. Add prometheus to the service manager.
3. Graceful shutdown of the services (including consensus).
4. Some code refactor regarding consensus graceful shutdown.
Co-authored-by: Rongjian Lan <rongjian.lan@gmail.com>
* Add ethereum transaction support
* fix fmt
* fix lint
* add new block body fork logic
* add back deleted code
* lower case block version const
* add eth txn into new version of block
* refactor back
* refactor more
* fix test code
* Fix build
* Fix build
* revert eth txn in block and block proposal
* fix build
* fix build
* fix build
* Add ethereum transaction support
* fix fmt
* fix lint
* add new block body fork logic
* add back deleted code
* lower case block version const
* add eth txn into new version of block
* refactor back
* refactor more
* fix test code
* Fix build
* Fix build
* [node] graceful shutdown validator and leader
when leader/validator is in commit phase, wait until it finishes the consensus
before shutdown the node.
This is useful for node upgrade to avoid unexpected blockchain state.
Signed-off-by: Leo Chen <leo@harmony.one>
* [node] stop block proposal service on leader
Signed-off-by: Leo Chen <leo@harmony.one>
* [node] warn on block proposal stopped
Signed-off-by: Leo Chen <leo@harmony.one>
Co-authored-by: Rongjian Lan <rongjian.lan@gmail.com>
[viewchange] encapsulate view ID in the State struct
do NOT ues consensus.current directory to set/get viewID
use the following functions
consensus.SetCurViewID, consensus.SetViewChangingID
consensus.GetCurViewID, consensus.GetViewChangingID
Signed-off-by: Leo Chen <leo@harmony.one>
* Optimize crosslink verification logic and add more cache (#3032)
* make crosslink broadcast smarter and more efficient (#3036)
* adjust catch up speed to be a sane number
* Fix crosslink broadcast condition (#3041)
* processing staking and normal transactions together
* bug fix and removing unwanted method
* made sense to rename and move the TransactionsByPriceAndNonce to pool type
* accounting for prestaking blocks, need to keep the previous order for processing txs
* Clean up logs
* remove log
* Refactor block period logic; wait for 100% sig before timeout
* logic fixes
* optimize now init
* always wait for 2s for grace period
* More logs and checks before processing staking txns
* fix import
* Refactor block proposal
* Various fixes and todos for auditing; add log for collect rewards
* Fix lint
* fix comment
* [staking][validation][protocol] Limit max bls keys
* [staking-era] Fold banned and active into single field
* [slash][effective] Remove LRU cache for slash, change .Active to enumeration
* [slash] Remove leftover wrong usage of Logger
* [slash][offchain] Only Decode if len > 0
* [offchain] cosmetic
* [slash] Remove some logs in proposal
* [webhook] Move webhook with call for when cannot commit block
* [shard] Finally make finding subcommittee by shardID an explicit error
* [node] Whitespace, prefer literal
* [webhook] Report bad block to webhook
* [slash] Expand verify, remove bad log usage, explicit error handle
* [slash] Check on key size
* [slash] Explicit upper bound of pending slashes
* [slash] Use right epoch snapshot, fail to verify if epoch wrong on beaconchain
* [multibls] Make max count allowed be 1/3 of external slots
* [quorum] Remove bad API of ShardIDProvider, factor out committee key as method of committee
* [verify] Begin factor out of common verification approach
* [project] Further remove RawJSON log, use proper epoch for snapshot
* [slash] Implement verification
* [slash] Implement BLS key verification of ballots
* [rpc] Keep validator information as meaningful as possible
* [staking] Never can stop being banned
* [slash] Comments and default Unknown case of eligibility
* [slash] Be explicit on what input values allowed when want to change EPOSStatus
* [consensus] Remove unneeded TODO
* [verify] Add proper error message
* [rpc] Give back to caller their wrong chain id
* [chain] Add extra map dump of delegation sizing for downstream analysis
* [engine] Less code, more methods
* [offchain] More leniency in handling slash bytes and delete from pending
* [validator] Remove errors on bad input for edit
* [webhook] Loop over all self address
* [rpc] Remove accidential recursion
* [rpc] Never ignore errors, fix RPC with current snap info, no div by zero
* [rpc] Expose BadBlock for downstream
* [availability] Thread through missing signers for offchain usage in WriteBlockWithState
* [project] Merge with master, accept master conflicts
* [offchain] Remove inactive signers from offchain data, should fix median-stake issues
* [rpc] Test idea that returning error cause jenkins failure
* [staking] Put editvalidator back into validator list cache
* [rpc] Add network voting-power to validator information
* [votepower] Extend roster with raw percentage
* [rpc] If have validator but not snapshot, give back validator
* [rpc] Expose effective-stake as well
* [project] Remove missedSignThreshold map, use instance for stakedslots count
* [staking] Remove edit-validator idea from previous filter idea
refactoring to multibls package to avoid cyclic dependency and better structure
fixing import issue
fixing more imports
fixing a bug that causes nil pointer for explorer node
fixing a bug in the super majority vote
changing config to have same epoch for cross link and staking
fixing the compilation error
addressing PR comments
leaving todo for assuming order between pub/pri keys pairs
fixing goimports
removing unwanted warning message
Rongjian Lan
Jacky Wang
Leo Chen
Ganesha Upadhyaya
Edgar Aroutiounian
flicker-harmony