* Add shard config in status message
* verify shard config for status message
* Notify peer connected to sync layer after status message exchanged
* Do not dial to shard config mismatched peers
* Upgrade network protocol version
* disconnect peer instead of ban peer if shard config mismatch
* Add python test for TCP connection by shard config
* Supports to randomly sync historical files
* Add name for random file sync batcher
* Remove sync store metrics since multiple random batcher created
* opt log
* ignore pruned or finalized historical file
* Add python tests for historical file sync
* Hotfix for python tests caused by unexpected file deletion
* add more info when launch blockchain node failed
* add stdout if blockchain launch failed
* seek stdout and err to 0 if failed to launch blockchain
* Improve zg chain port to avoid port conflict in parallel execution
* fix float issue
* Fix py failures
* Add new P2P protocol NewFile
* Publish NewFile message when any file finalized
* handle NewFile message in router
* handle NewFile in sync servic to write in db
* use propagation source to handle NewFile message
* Disable sequential sync and store new file in v2 sync store
* Add shard config in FindFile
* Add AnnounceFile RPC message in network layer
* do not propagate FindFile to whole network
* Mark peer connected if FileAnnouncement RPC message received
* fix unit test failures
* Change P2P protocol version
* Ignore py tests of sequential auto sync
* Add py test for auto sync v2
* fmt code
* remove dummy code in py test
* fix random test failure
* Add comments
* Enable file sync protocol v2 in config file by default
Currently we use the config `network_libp2p_nodes` to connect nodes
in the tests. This will not be retried, so if an early node starts
too slowly, other nodes may fail to connect to it.
* fix: pad tx based on the on-chain contract result.
* chore: update contracts
* test: shard submission
---------
Co-authored-by: Peilun Li <peilun.li@confluxnetwork.org>
* Use inner lock in storage.
* Remove mut.
* Remove async lock for storage.
* Fix tests and warnings.
* Use spawn_blocking for storage task.
* Fix clippy.
* Finalize the new tx at last.
* Revert "Finalize the new tx at last."
This reverts commit b56ad5582d.
* Wait for old same-root txs to finalize.
* Use async storage in miner.
* Update rust version to 1.79.0.
* Use Vec to avoid stack overflow.
* Fix unused warning.
* Fix clippy.
* Fix test warning.
* Fix test.
* fmt.
* Use async storage in pruner.
* nit.
* Save test logs.
* Prepare to handle shard config in chunk pool.
* Allow files to be finalized with only sharded data.
* Handle shard config change.
* Only sync needed data in shard config.
* Fix clippy and tests.
* Implement Pruner.
* Put pruner in a crate.
* Fix clippy.
* Add rpc zgs_getShardConfig.
* Fix.
* Increase wait time.
* Add pruner_test and use max_num_chunks instead of size_limit.
* Store shard config for peers and choose sync peers accordingly.
* Add test and fix sync.
* Fix clippy and test.
* Fix some ut.
* Add AnnounceShardConfig gossip and fix tests.
* Add sharded tx finalize check in LogManager.
* Try,
* Rename.
* Longer timeout for mine_test.
* Save test logs.
* Change PoraHash compute
* Change padSeed compute
* Refactor
* Support mining on sharded storage
* Detect single core performance and set correct params for test
* Fix clippy
* Fix an overflow bug
* Change miner id logic & request miner id automatically
* Not enable all features in the test workflow.
* Auto configurable mining period
* Adjust test params for ci
* Separate script to build blockchain binaries
* auto remove http_proxy env if configured
* build bsc binary in run all process
* auto install binary when run a single test
* Add necessary output when build or download binary
* create tmp dir if absent