Adds integration tests to catch bug where the mark storage was being cleared before a mark began. Instead, the mark storage is now cleared upon preparing the mark sweep pruner
Fixes bug where the pending marks where not being checked while pruning was occurring. By removing the flush in sweepBefore, the existing tests catch the bug.
Signed-off-by: Ratan Rai Sur <ratan.r.sur@gmail.com>
Default logging should be 'null' - which means that Besu will do nothing
to the Log4J2 logging levels and rely on the built in or user supplied
log4j2.xml configuration files to set up logging.
Signed-off-by: Danno Ferrin <danno.ferrin@gmail.com>
* Update version to 1.2.5-SNAPSHOT (#42)
Signed-off-by: Edward Evans <edward.joshua.evans@gmail.com>
Signed-off-by: Abdelhamid Bakhta <abdelhamid.bakhta@consensys.net>
* introduce `VirtualOperation`
- add `isVirtualOperation` method to `Operation` interface: determines whether or not the operation has been virtually added to the contract code.
- default implementation of `isVirtualOperation` returns `false`
- create `VirtualOperation` that wraps a real `Operation` and returns `true` for `isVirtualOperation` method.
- change `constructor` of `EVM` to pass a `GasCalculator` used to create the `endOfScriptStop` and `invalidOperation` fields.
- change visibility of `EVM.operationAtOffset` to `package` for testing purpose. (add `VisibleForTesting` annotation)
- create `EVMTest` to test `operationAtOffset` method .
Signed-off-by: Abdelhamid Bakhta <abdelhamid.bakhta@consensys.net>
* Update version to 1.2.5-SNAPSHOT (#42)
Signed-off-by: Edward Evans <edward.joshua.evans@gmail.com>
Signed-off-by: Abdelhamid Bakhta <abdelhamid.bakhta@consensys.net>
* Store db metadata file in the root data directory
The database metadata file should be stored in the root data directory rather than the database subdirectory.
The database subdirectory is owned by the database itself and should not be directly manipulated by the node.
- first look in the data directory for the metadata file
- if the metadata file is found there, process it as normal
- if no metadata file is found in the root directory, look in the database subdirectory
- if the file is found here, copy it to the root directory, and run based on the root directory version
Signed-off-by: Abdelhamid Bakhta <abdelhamid.bakhta@consensys.net>
* add logs
Signed-off-by: Abdelhamid Bakhta <abdelhamid.bakhta@consensys.net>
* create database directory if database not detected
Signed-off-by: Abdelhamid Bakhta <abdelhamid.bakhta@consensys.net>
* change plugin API know hash
Signed-off-by: Abdelhamid Bakhta <abdelhamid.bakhta@consensys.net>
Some of the logging levels produce surprising results, for example the
call to --help spits out info about rocksdb starting, which we don't
need. Turn down plugin start/stop to debug and using an existing DB to
debug. Config options are only developer relevant, so down to trace.
Signed-off-by: Danno Ferrin danno.ferrin@gmail.com
* Only publish on Azure for SNAPSHOT releases
Signed-off-by: Edward Evans <edward.joshua.evans@gmail.com>
* Fix comment char
Signed-off-by: Edward Evans <edward.joshua.evans@gmail.com>
* adding in spdx-license-identifier & updated check for the same; removing license check from spotless
Signed-off-by: Joshua Fernandes <joshua.fernandes@consensys.net>
* Change CheckSpdxHeader to a task.
Signed-off-by: Adrian Sutton <adrian.sutton@consensys.net>
Expose EthPeer validation state so that the Synchronizer can choose peers based on whether or not they have been fully validated. This allows us to use only fully validated peers when choosing a pivot block.
Signed-off-by: Meredith Baxter <meredith.baxter@consensys.net>
Signed-off-by: Danno Ferrin <danno.ferrin@gmail.com>
Update discovery logic to consider a peer with an unknown discovery endpoint to be unknown regardless of whether we've encountered a peer with the same node id before. This makes the discovery logic more forgiving in the face of node restarts.
If nodeA bonds with nodeB, then nodeB leaves the network and later comes back with a different ip address or listening port, nodeA would previously continue trying to communicate with nodeB at its original address. With these changes, nodeA will now treat the restarted nodeB as a new peer and communicate with it on its updated endpoint. Additionally, nodeB's information will be updated in the peer table so that neighbors requests return updated information on this node.
Signed-off-by: Danno Ferrin <danno.ferrin@gmail.com>
Change the retesteth port to 47710 (0xba5e) so
(a) it does not conflict with other clients when started w/o args
(b) it is flavorful to the "Besu" name
Signed-off-by: Danno Ferrin <danno.ferrin@consensys.net>
This allows the acceptance testing DSL to be published as part of the
Besu internal jar files - which in turn means the DSL can be reused
in other projects.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
* updating jenkinsfile to skip publishing snapshots to bintray
* updating docs link in readme
Signed-off-by: Joshua Fernandes <joshua.fernandes@consensys.net>
Move header validations that extract the signer key out of the "light"
validation mode. Reduces fast sync time on goerli 75%
Signed-off-by: Adrian Sutton <adrian.sutton@consensys.net>