* CORDA-3018 Allow loading an untrusted contract jar if another attachment exists that was signed with the same keys and uploaded by a trusted uploader
`TransactionUtils.isAttachmentTrusted` requirements have been relaxed
to allow an untrusted attachment to be loaded as long as another
attachment exists that is signed by the same keys and was uploaded
by a trusted uploader.
The requirement of containing the same contract classes has been
removed. Therefore the contents of the existing trusted attachment
no longer matters.
* CORDA-3018 Allow a subset/intersection of signers in `isAttachmentTrusted`
Allow a subset/intersection of signers to satisfy the signer
requirements of `isAttachmentTrusted`. This allows an "untrusted"
attachment that is signed by one or more keys to be "trusted" as long
as another trusted attachment already exists that is signed by at least
one of the "untrusted" attachments signers.
A cache of trusted and untrusted public keys is now held (replacing the
previous cache of `List<PublicKey>`.
Tests have been added to `NodeAttachmentServiceTest` to confirm that
an attachment query using an `EQUAL` statement will actually return
attachments that are signed by any of the keys passed into the query.
Confirming this allowed an `EQUAL` query to satisfy the search that
had to be done as part of this change.
`MockAttachmentStorage`'s query criteria was updated to better match
the real `NodeAttachmentService` implementation.
* CORDA-3018 Update cache name and kdoc on `isAttachmentTrusted`
* CORDA-3018 Verify that chains of trust do not occur
* CORDA-3018 Switch keys around to improve chain of trust tests
* [ENT-3801] Store transactions in the database during transaction resolution (#2305)
* ENT-3801: Store downloaded txns as part of the backchain resolution into the db rather than the checkpoint
It's very inefficient to store the downloaded backchain in the checkpoint as more of it downloaded. Instead, if a threshold is reached (which currently defaults at 0) then the backchain is stored in the transactions table as unverified. A new is_verified column has been added to track this. Initially testing on the OS codebase has been very promising but unfortunately this current code is not quite ready. I had to quickly port it to ENT as this is meant to be an ENT-only optimisation.
To that effect, there is a TransactionResolver abstraction with two implementations: an in-memory one which has the old behaviour, and which will be the behaviour for OS, and a db one.
DBTransactionStorage hasn't been fully updated and I had to comment out the optimistic path for now.
Most of these changes will need to be ported to OS to keep the merge conflicts in check, but obviously not DbTransactionsResolver and the "is_verified" changes in DBTransactionStorage. DBTransactionStorage does have other refactoring which will make sense to port though.
* [ENT-3801] Start work on allowing modifications in AppendOnlyPersistentMap
* [ENT-3801] Add transaction resolver tests
* [ENT-3801] Adjust suspendable annotations
* [ENT-3801] Fix the ResolveTransactionFlow tests
* [ENT-3801] Update ResolveTransactionsFlow tests
* [ENT-3801] Add a liquibase migration script for isVerified
* [ENT-3801] Ensure the migration runs in the correct place
* [ENT-3801] Handle resolution of already present transactions
* [ENT-3801] Fix compile error in performance test app
* [ENT-3801] Logging and comment updates, plus a test case
* [ENT-3801] Add a notary change resolution test
* [ENT-3801] Add a contract upgrade transaction test
* [ENT-3801] Change new column to be a character based status
* [ENT-3801] Migration script type change
* [ENT-3801] Address first round of review comments
* [ENT-3801] Update variable names in AppendOnlyPersistentMap
* [ENT-3801] Another variable name clarification
* [ENT-3801] Fix missing name changes
* [ENT-3801] Make the signature list immutable when constructing cache value
* [ENT-3801] Add a locking strategy for unverified transactions
* [ENT-3801] Address tidying up review comments
* [ENT-3801] First attempt at ensuring locks are released after commit
* [ENT-3801] Remove references to old cache name
* [ENT-3801] Update locking logic
* [ENT-3801] Fix potential deadlock with read/write transaction locks
* [ENT-3801] Remove read locks, and ensure minimal extra suspends
* [ENT-3801] Fix build issues in tests
* [ENT-3801] Use the correct clock when calculating sleep durations
* [ENT-3801] Add a pessimism flag for writing verified transactions
* [ENT-3801] Change logging statement to debug
(cherry picked from commit 8ab6a55e17)
* [NOTICK] Fix up imports for some changed files
* [NOTICK] Fix transaction resolution tests
* [NOTICK] Reinstate the DBTransactionsResolver
* [NOTICK] Add the topological sort back to recordTransactions
* [NOTICK] Adjust test case to remove dependency on query ordering
* [NOTICK] Make test code match that in ENT
* CORDA-2617: Add failover listeners to terminate node process
This is a backport of changes done in Corda Enterprise.
It will be triggered in case of:
a) Loss of connectivity to in-built Artemis for Bridge Control;
b) Loss of connectivity to in-built Artemis for P2P connectivity.
Note on merge to CE: Disregard these changes and take whatever CE already has.
* CORDA-2617: Update documentation on stability of Corda Node
* CORDA-2617: Documentation update after discussion with @mnesbit
* parallelize download of nodeInfos
* actually call new list based addNodes method
* address review comments
fix NetworkMapUpdaterTest
* ensure threadpools are shutdown after network-map download is completed
* use NamedThreadFactory instead of re-implementing it.
* fix imports after rebase
* address review comments
* remove extra whitespace
* Initial commit based on experimental kryo hook agent.
* WIP
* Added documentation.
* Additional improvements and documentation following more testing.
* Added field level instrumentation + basic type handlers for String, byteArray, charArray, primitive types.
* Working version (without array type handling)
* Missing build.gradle file.
* Handle display of Arrays and String.
Pruning output to avoid repetition (by loop depth, object count).
* Added configurable StackDepth (for display purposes) and filter out ProgressTracker stacks.
* Further array handling (Object arrays, 2D, 3D), improved display and general code cleanup.
* Various fixes and improvements following demo to RP.
* Clean-up
* Further clean-up
* Set checkpoint id before deserialization.
* Update documentation
* Final clean-up.
* Minor documentation fixes.
* Updates following PR review feedback.
* Add changelog entry.
Change to `SingleThreadedStateMachineManager`.
Instead of throwing an exception when a flow's fiber cannot be found,
just log the message at info level.
(cherry picked from commit 70b2a94fda)
Allow Hash Constraints to propagate to Signature Constraints. When
propagating, the new jar is added to the transaction instead of the
original contract attachment (jar).
The following requirements must be met to do so:
- System property "net.corda.node.disableHashConstraints" must be set
to `true`
- The min platform version must be 4 or greater
- There must be an input state with a hash constraint
- There must be no output states with a hash constraint
- The new jar must be signed
If these requirements are not met, the original Hash
Constraint will be kept and the contract attachment related to it is
also used.
This transition is done at the start of `handleContract` as it is not
the normal path a transition would follow. It is considered a backdoor
and should be treated separately from the rest of the attachment and
constraint logic. Furthermore, it will only work in private network
since all nodes must set the special `disableHashConstraints` flag.
* Fix test execution flakiness on fast hardware: set node info polling interval to 1 second in DriverDSL Node Startup (was only being set to 1 sec in Node Registration). Follow-up to https://github.com/corda/corda/pull/5240
* Diagnostics: add additional logging to NodeInfoFilersCopier.
* Diagnostics: add additional logging to NodeInfoWatcher
* Downgrade logging level severity to debug.
* Make atomic counter unique across instances.
Change to `SingleThreadedStateMachineManager`.
Instead of throwing an exception when a flow's fiber cannot be found,
just log the message at info level.
When set to true the RPC client will:
* automatically reconnect when the connection is broken
* simple RPC calls will block until connection is established
* Observables returned from RPC will automatically resubscribe on reconnect so the client continues to receive events. This doesn't guarantee that events will not be lost during the reconnect.
As a general purpose API, allPersisted should not be loading the entire contents of the database table into memory. Instead now it returns a Stream for processing of elements.
* CORDA-3028: Add Node Diagnostics Info RPC Call - Backport a diff from Corda ENT
* CORDA-3028: Add Node Diagnostics Info RPC Call - Rename a value for consistency
* CORDA-3028: Add Node Diagnostics Info RPC Call - One more change
* Moved core.flows and core.contract tests
* move internal tests
* move node, schemas and serialization tests
* Move transactions tests
* Move test access helper into test package
* Move remaining tests, and create access helpers to test internals.
* Fix test dependencies.
* Delete some cruft that flagged up on the API checker
* Add readme explaing where tests need to go.
* fix typo
* add a shared memory port allocator to allow multiple processes to share a single allocation pool
* remove dangerous reset function on port allocator
* set forkCount = 2 in node integration test
* only allow one build of a cordapp at any given time for Driver tests
* make all portallocation requests use same starting point
* globally set forks to 6
* tweak forking parameters to allow parallel builds
* tweak unit test parallelism
* 2 workers for integrationTest
* some more tweaks for parallel builds
* some more tweaks for parallel builds
* seems that 49K is not the start of ephemeral ports on all kernels
* tweak parallel settings
* try fix RPC shutdown test in parallel env
* add some logging for RPC shutdown test
* added some logging around PortAllocation tests - try figure out where they are getting stuck
* added some logging around PortAllocation tests - try figure out where they are getting stuck
* fix api-scanner tests
* minimize api changes
* revert to complying with existing API
* add the AtomicInteger for api compatibility reasons
* make sizing script executable
* address review comments pt1
* address review comments pt2
* fix compile errors after review comments
* return to using home dir as temp dir seemed to interact badly with gradle
* CORDA-3021: Introduce `SignOnlyCryptoService` and use it whenever possible
Also modify `CryptoServiceFactory` to show how sign only implementation can be created.
* CORDA-3021: Undo some of my earlier changes
Which after discussion with @dimosr and @fowlerrr proven to be contradictory.
Dumps all the node's checkpoints as JSON into a single zip file in the node's directory. The output contains:
* All the fields for the top-level flow
* The current sub-flow call stack, along with the current progress tracker step for each sub-flow
* The event that suspended the flow, which if it's a send or sendAndReceive will show the payload that was sent
* Low level information on the active sessions with other peers
* Initial commit with a test that can detect a logger that isn't initialised with by lazy that gets called during startup.
* Test improvement. Test fix for api stability.
* Added explanatory comment for the test.
* Changes according to PR review. Added empty file check to make test more robust.
Party and AnonymousParty have been retrofitted to implement this interface and are currently the only supported types. A new FlowLogic.initiateFlow(Destination) method overload to easily support the addition of new destination types in future versions.
To prevent making `dumpCheckpoints` part of the public API a new
interface, `InternalCordaRPCOps` has been created and the function
has been moved there. `InternalCordaRPCOps` inherits from
`CordaRPCOps`.
`CordaRPCOpsImpl` now implements `InternalCordaRPCOps`.
`RunShellCommand` and `StringToMethodCallParser` required additional
changes due to issues handling inherited functions. This has only been
raised now due to `InternalCordaRPCOps` inheriting from `CordaRPCOps`.
Many classes have had references to `CordaRPCOps` changed to
`InternalCordaRPCOps`.
* ENT-3642: move the crypto service builder method to node-api
* ENT-3642: add arg for different crypto services
* ENT-3642: add arg for cryptoservice config
* ENT-3642: add null check for BCC requirement
* ENT-3642: remove double-bang operator
* ENT-3642: cryptoservice factory method moved to its own class, improve API
* ENT-3642: remove import
Store the flow dump zip in the logs directory.
Only allow one dump to be created at a time. This is protected by
using a `AtomicInteger` lock and checking for the zip's existence.
* Revert usage of Gradle 5 useJUnitPlatform() which causes significant test execution performance degradation.
* Remove completely top-level Java CordaCaplet unit tests causing consistent TC failure on Azure VMs:
java.security.NoSuchAlgorithmException: no such algorithm: NONEwithEdDSA for provider BC
* Added periodic message logged from a timer which is switched on and off by the node drain events.
* Timer also spawns on start() to ensure that if a node is shutdown in draining mode and started with it, it will continue logging messages.
* Reduce test execution times by explicitly configure quasar package exclusions (using new quasar util extension introduced in Corda gradle plugins 5.0.1-SNAPSHOT).
* Remove SNAPSHOT from Corda gradle plugins version identifier.
* Incorporating PR review feedback from CR.
* Minor consolidation clean-up.
* Minor consolidation clean-up.
* Minor consolidation clean-up.
Dumps all the node's checkpoints as JSON into a single zip file in the node's directory. The output contains:
* All the fields for the top-level flow
* The current sub-flow call stack, along with the current progress tracker step for each sub-flow
* The event that suspended the flow, which if it's a send or sendAndReceive will show the payload that was sent
* Low level information on the active sessions with other peers
* CORDA-2838 Set Artemis memory config.
* CORDA-2838 Cannot have page size larger than max size.
* CORDA-2838 Use real slow consumers. Need to see if the old config settings can work with a global limit to avoid this.
* What went wrong:
Caused by: org.gradle.api.GradleException: failed to read class file /Users/josecoll/IdeaProjects/corda-jdk11/node/build/classes/java/test/net/corda/node/services/events/FlowLogicRefFromJavaTest$JavaNoArgFlowLogic.class
* Upgrade gradle wrapper to Gradle 5.0
* Upgrade to use locally deployed version of Capsule plugin (using Gradle 5.0)
* Upgrade to use Corda Gradle Plugins 5.x (inclusive of Gradle 5.0 compatibility fixes)
* Fix compile-time problems resolving log4j packages.
* Update to use Artifactory deployed version of Capsule plugin (using Gradle 5.0)
* Upgrade to use Gradle 4.2.1 (excluding default memory setting change: command line client now starts with 64MB of heap instead of 1GB.)
* Added explicit compile-time dependency on 'de.jensd:fontawesomefx-commons:8.13' (compile-time error in tools:explorer without)
* Update "wrapper" gradleVersion to 5.4.1
* Update Capsule plugin version label to reference R3 forked build.
* Commit all gradle upgrade changes following: ./gradlew wrapper --gradle-version 5.4.1
* Increase maximum heap memory for Test tasks to 1GB, and upgrade build-scan plugin to 2.2.1.
* Increase Test tasks' maximum memory to 1.5GB - what could go wrong?
* Up maxHeapSize to 2g for gradle test runner (global setting).
* Added explicit compile-time dependency on 'de.jensd:fontawesomefx-commons:8.13' (compile-time error in tools:demobench without)
* Added forkEvery for node Unit tests.
* Up :node test task 'forkEvery' to 10.
* TC test execution tuning (:core, :node => forkEvery 10, default JVM heap size)
* TC test execution tuning (bump JVM heap size up to 1g for :node)
* TC test execution tuning (bump JVM heap size up to 1.5g for :node)
* TC test execution tuning (re-instate global JVM heap size of 1Gb)
* TC test execution tuning (re-instate JVM heap size of 2Gb for :node)
* Update Corda Gradle plugins to 5.0.0
* Updated plugin resolution order + renamed artifactory URL to use "software.r3.com"
* Reorder plugin resolution such that mavenlocal() is always queried first.
This allows a different signed version of the same CorDapp to be automatically trusted.
This reverts "[CORDA-2575] Allow users to whitelist attachments by public key config (#5035)"
When a `UnexpectedFlowEndException` or a `FlowException` is received the
peer that the exception was thrown from will be added to the stacktrace.
This is due to it being easier to see and a field that developers
are much less likely to override.
A nullable field `peer` has been added to `FlowException` and
`UnexpectedFlowEndException`. This is read later on (when peer info
is not available) to append the peer info to the stacktrace.
* Initial version of new(old) dialog that won't print a stacktrace for rpc exceptions.
* Decoupled CordaVersionProvider. Moved common files to common-logging to lower dependencies on the node explorer.
* Removed unused import and duplicate documentation comment.
* Moved error code rewrite policy in the new common/logging module according to PR review.
* Removed extra line.
* Updated log4j configurations with new package name where logging policies will be contained.
* Included common-logging module with cliutils.
* setting non existant base directory now yields suitable error message
* corrections to missing base directory message and comments added for clarity
* added check for valid base directory before node loggin and bootstrapping
* removed uneeded import
* quantum computer joke back by popular demand and not back at the same time
A test was failing due to `assumeFalse` (to stop a test running on windows)
being inside of a `assertThatExceptionOfType` block. The exception it
threw was consumed by the block instead of junit.
If a single whitelisted constraint is being used by input states and the version of the cordapp changes + is signed, then the constraint will transition to a signature constraint.
When a `UnexpectedFlowEndException` or a `FlowException` is received the
peer that the exception was thrown from will be added to the stacktrace.
This is due to it being easier to see and a field that developers
are much less likely to override.
A nullable field `peer` has been added to `FlowException` and
`UnexpectedFlowEndException`. This is read later on (when peer info
is not available) to append the peer info to the stacktrace.
`ExceptionSerialisingRpcOpsProxy` was removing information about the
original exception in dev mode. This code has been removed. Although
there is no check in `ExceptionSerialisingRpcOpsProxy` for dev mode
(and also due to it being about serialising),
`ExceptionMaskingRpcOpsProxy` will handle the removal of exception data
in non dev mode (production mode).
* [CORDA-2522] Improve error reporting around failed flows (#5000)
* Improve error reporting around failed flows
* Fix an index to start from 1 when printed
* Address first set of review comments
(cherry picked from commit 24699cd7f4)
* [CORDA-2522] Follow up changes to error reporting around failed flows (#5006)
* Follow up changes to error reporting around failed flows
* Have FinalityDoctor report stack trace
* Revert changes to the DumpHistoryOnErrorInterceptor
(cherry picked from commit 2da597a5b7)
* Follow up changes to error reporting around failed flows
* Have FinalityDoctor report stack trace
* Revert changes to the DumpHistoryOnErrorInterceptor
* Initial version of new(old) dialog that won't print a stacktrace for rpc exceptions.
* Decoupled CordaVersionProvider. Moved common files to common-logging to lower dependencies on the node explorer.
* Removed unused import and duplicate documentation comment.
* Moved error code rewrite policy in the new common/logging module according to PR review.
* Removed extra line.
* Updated log4j configurations with new package name where logging policies will be contained.
* Included common-logging module with cliutils.
* setting non existant base directory now yields suitable error message
* corrections to missing base directory message and comments added for clarity
* added check for valid base directory before node loggin and bootstrapping
* removed uneeded import
* quantum computer joke back by popular demand and not back at the same time
* CORDA-2813 Add DbException handler to terminate process when the DB goes away or gives up.
Also add a helper to terminate the node and use it instead of calling Runtime.halt() in random places.
* CORDA-2813 Address threading/blocking concerns raised during review
* Tests for custom registry restrictions
* ENT-3121 restrict custom serialisation
* Remove redundant code
* Only count declared annotations
* Check annotation on superclasses, remove annotation from ByteArray
* Forbid custom serialization of primitive types
* Remove @CordaSerializable from another class that is always handled by custom serialisation
* Add log warnings to aid diagnosis of custom serialization issues
* Remove another annotation
* Remove another annotation
* Remove another annotation
* Remove another annotation
* Fixup api-current
* Fixup api-current
* KDocs on exceptions
Addressed review comments - removed nested if and added parameter name for the usage of requiresCertificates.
Removed unnecessary import and white space.
Normally, these requests are left unacknowledged in the MQ broker to allow recovery via installing the missing CorDapp. If this is not applicable then the request be dropped (and the other side informed of the error).
While the node is starting up, we now check for the presence of the certificates directory. This allows us to print out an easily understandable error message if the directory is not present. An exception is made for devMode, as devMode will result in the directory being created in any case.
* ENT-3057: Log hibernate warns and errors in different log
If a hibernate error occurs (deadlock, for example) that would cause a flow to be sent to the hospital, hibernate logs the warnings and errors before we do. This results in duplication in the logs, and pollutes the log. To solve this, we create a new log appender named diagnostic-{node-name}.log and log any org.hibernate messages of warn and above to that file. This way, messages are not lost, which means that the information can be retrieved if need be.
* Corrected indentation of comment (changed tab to space)
* Updated node-administration document to mention diagnostic logging change
* Fixed integration test. It was breaking because it was fetching the first log file in the folder, assuming there would be only one. This assumption is now invalid because the diagnostic log file that was introduced. Two tests were found that used similar logic to find a log file to examine, hence both were corrected to look for log files beginning with "node"
* Updated documentation as per review comments.
Add a cache for notary identities in the PersistentIdentityService. This
solves a reported problem where notary identity lookup fails if its network
map entry is missing, which results in an exception when trying to insert
a state into the vault after recording a transaction.
Test if we have access to the logging path (baseDirectory/logs) before attempting to write to them. This allows us to shut down gracefully with an easily understandable error message.
Without doing this, the log4j2 will attempt to access the logPath when it first uses the logger in the call() function.
