The Hash of a qbft payload is calculated by hashing the RLP'd bytes of the message-id, followed by the payload's encoded bytes.
RLP(msgCode, RLP(payload.encodedBytes()))
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
We were creating new in-memory storage segments each time we were
supposed to be retrieving it which prevented me from being able to test
what keys ended up being stored in that segment.
Signed-off-by: Ratan Rai Sur <ratan.r.sur@gmail.com>
* Decouple PoW from ethash
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* Address code review comments, create a dev network for ecip1049, prepare for keccak hasher
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* Add PoW function and a few simple tests as test vectors
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* Make the PoWHasher hash function a bit easier to understand
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* simplify and call out the code of the keccak hash function
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* support fixed difficulty for keccak mining
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* Fix the dev network config
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* Add comment to KeccakHasher
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* Increase fixed difficulty for the ecip1049 dev network to produce hashes a bit less often
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* spotless
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* fix test expectations
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* Fix javadoc issue
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* add acceptance test using keccak mining
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* add changelog entry
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
* Address code review comments
Signed-off-by: Antoine Toulme <antoine@lunar-ocean.com>
This is the first step in supporting switchable consensus mechanisms.
Specifically this allows additional protocol specs to be inserted to the protocol schedule at milestones other than that explicitly specified in the genesis config.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
This allows Besu to interop with an Quorum IBFT-1 network which has been updated to use a 2/3 validator threshold, rather than 2F+1.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
When interop testing between Besu and Quorum, it was found
that Quorum's management of the ibft protocol had somewhat
changed, and resulted in besu failing to peer with Quorum
To overcome this, the version of the istanbul protocol in
Besu was upgraded to 99 (matching Quorum's).
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
QBFT no longer validates all fields of the block header (eg nonce mixhash), as these have no bearing on the
safety model of the protocol.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
Add QBFT rpc methods and namespace. At the moment, they are copied verbatim from IBFT namespace and may change in future once specifications are finalized.
Signed-off-by: Usman Saleem <usman@usmans.info>
* Enable Besu to import blocks containing quorum style private transactions
* Add RPC to accept quorum style raw private transactions
Signed-off-by: Stefan Pingel <stefan.pingel@consensys.net>
Signed-off-by: Lucas Saldanha <lucascrsaldanha@gmail.com>
Co-authored-by: Lucas Saldanha <lucascrsaldanha@gmail.com>
All received QBFT messages are now validated with respect to protocol safety prior to be accepted to QBFT business logic.
This includes updates to the QBFT integration test framework.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
* Finish internal ConstantinopleFix->Petersburg transition, including
class names. Only backwards compatibility and retesteth use of the
name will remain.
* Remove four unused config fields from json configs. These are fields
with no code uses of any sort. All are implied by other fields.
eip150Hash, eip155Block, eip160Block, and daoForkSupport.
* Remove redundant fork block fields where their value is implied
Signed-off-by: Danno Ferrin <danno.ferrin@gmail.com>
This supplies the necessary validation logic for parsing received RoundChange messages.
This logic is wired into the QBFT business logic.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
The code required to deserialise signed payloads for QBFT is verbose and can be reduced by templating.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
Creates the validator for qbft commit messages, but does not wire it into business logic at this stage.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
Adds the Prepare validator, but does not wire it into the QBFT message validation at this stage.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
Move the messageFactory out of IbftFinalState such that the IbftFinalState can be reused between IBFT and QBFT.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
This adds the proposal and roundchange messages required for QBFT.
This does not wire in the ability to transmit/receive these messages, it just defines the message structure.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
Aspects of the consensus mechanism associated with block creation, and validation have been moved
from the IBFT package, into into consensus/common, such that they can be reused for the QBFT
implementation.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
This adds the Prepare and Commit messages used in the QBFT protocol, this is an incremental movement toward QBFT - and does not offer new functionality.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
When validating a RoundChange message, its important that not only
are there sufficient Prepare msgs in the PreparedCertificate, but
also that each are from a different validator (i.e. cannot simply
count the number msgs, but must count distinct authors).
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
Move the block header validation errors logging levels from trace and
debug to info. Also, include a standard prefix "Invalid block header: "
in each of the log lines.
Signed-off-by: Danno Ferrin <danno.ferrin@gmail.com>
Since the EIP-1559 transition is going away, simply use a set of
accepted transactions for transaction validator that we'll check
against.
Don't assume that there are two transaction types and instead check
what type the transactions are.
Use guessType when we're deserializing from json.
Signed-off-by: Ratan Rai Sur <ratan.r.sur@gmail.com>
Moves IBFT events to the common package for reuse with the QBFT implementation.
This necessitated the creation of BftEventHandler interface, such that a generic IbftController can be passed around, rather than the literal IbftController.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
Moving the IbftMessage (and rename to BftMessage) to the common package such that it can be reused as part of the QBFT implementation.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
SignedData is to be used as part of QBFT, and as such the IBFT specific aspects of this class are to be moved into a separate class (allowed SignedData to be moved into the common package)
Signed-off-by: Trent Mohay trent.mohay@consensys.net
This is the first step in making aspects of the IBFT2 code base common, such that it can be reused as part of the QBFT implementation which is coming shortly.
Signed-off-by: Trent Mohay trent.mohay@consensys.net
This is a stronger check than in the past, which potentially allowed for a the IBFT system to handle messages for the current round, even if a block had already been imported (but had not yet received the new-block event) - which in turn, could lead to a network fork in a adversarial environment.
Signed-off-by: Trent Mohay <trent.mohay@consensys.net>
For every block produced and imported directly log the same information
we would get from other import paths.
* If this node produced the block, note that with "Produced" instead of
"Imported"
* Log the size of the pending transaction pool
* Only log IBFT rounds at INFO if they go past round 0, DEBUG otherwise.
Signed-off-by: Danno Ferrin <danno.ferrin@gmail.com>
Trent Mohay
Ratan (Rai) Sur
Antoine Toulme
Daniel Lehrner
Usman Saleem
Jason Frame
Trent Mohay
Stefan Pingel
Danno Ferrin
David Mechler