Infinispan

Clone Tools
  • last updated a few minutes ago
Constraints
Constraints: committers
 
Constraints: files
Constraints: dates
5.2.3.Final

  1. … 27 more files in changeset.
ISPN-2787 NPE after ReplaceCommand

Assert the expected exception is thrown in MergeDuringReplaceTest

ISPN-2787 NPE after ReplaceCommand

Assert the expected exception is thrown in MergeDuringReplaceTest

ISPN-2787 NPE after ReplaceCommand

* Add MergeDuringReplaceTest to reproduce the issue.

* Fix one of the ControlledRpcManager.invokeRemotely methods that was ignoring the actual result and was always returning null.

* Minor logging improvements to be able to debug this issue.

* Remove harmfull optimization from RpcManagerImpl.invokeRemotely(Collection<Address> recipients, ReplicableCommand rpcCommand, ResponseMode mode, long timeout, boolean usePriorityQueue, ResponseFilter responseFilter) that was not going remotely and neither threw any exception if the destinations were no longer in cluster and current node was the only member.

ISPN-2884 - Upgrade to JGroups 3.3.0.Beta1

ISPN-2889 AbstractIterator shouldn't allocate a larger result array than is necessary

ISPN-2865 Error calling clear() via JMX on a transactional cache with autoCommit=false

ISPN-2865 Error calling clear() via JMX on a transactional cache with autoCommit=false

ISPN-2888 ManagedOperation.name() is ignored when registering JMX operations

ISPN-2888 ManagedOperation.name() is ignored when registering JMX operations

Clean up query iterators (encapsulate fields in AbstractIterator, etc.)

ISPN-2825 ClusterTopologyManagerImpl should not hold a lock while invoking an RPC

If the REBALANCE_START command takes a long time to be processed on

one of the nodes, the REBALANCE_CONFIRM commands from the rest of

the nodes are blocked and end up filling the OOB thread pool.

Add two new tests for large clusters.

Move CacheCreationStressTest to the profiling group.

ISPN-2825 ClusterTopologyManagerImpl should not hold a lock while invoking an RPC

If the REBALANCE_START command takes a long time to be processed on

one of the nodes, the REBALANCE_CONFIRM commands from the rest of

the nodes are blocked and end up filling the OOB thread pool.

Add two new tests for large clusters.

Move CacheCreationStressTest to the profiling group.

ISPN-2885 Expose parameter names in MBean metadata ISPN-2883 Use special naming to overcome JMX method overloading with RHQ operation name uniqueness

  1. … 3 more files in changeset.
ISPN-2885 Expose parameter names in MBean metadata ISPN-2883 Use special naming to overcome JMX method overloading with RHQ operation name uniqueness

  1. … 3 more files in changeset.
ISPN-2635 and ISPN-2536: Support total order based commit protocol

  1. … 140 more files in changeset.
ISPN-2787 NPE after ReplaceCommand

* Add MergeDuringReplaceTest to reproduce the issue.

* Fix one of the ControlledRpcManager.invokeRemotely methods that was ignoring the actual result and was always returning null.

* Minor logging improvements to be able to debug this issue.

* Remove harmfull optimization from RpcManagerImpl.invokeRemotely(Collection<Address> recipients, ReplicableCommand rpcCommand, ResponseMode mode, long timeout, boolean usePriorityQueue, ResponseFilter responseFilter) that was not going remotely and neither threw any exception if the destinations were no longer in cluster and current node was the only member.

ISPN-2881 - org.infinispan.loaders.jdbc.mixed.JdbcMixedCacheStoreTest2 - Cache store tests expect purgeSynchronously to be enabled * Added the purgeSynchronously parameter * Renamed the Class names so that the maven includes running them in testsuite.

ISPN-2509 Force transport instance creation if pingOnStartup enabled

ISPN-5013 JSR-223 Scripting

    • -0
    • +10
    /client/hotrod-client/src/test/resources/test.js
  1. … 58 more files in changeset.
ISPN-2878 Default iteration on QueryCache is changed to EAGER iteration

ISPN-2878 Default iteration on QueryCache is changed to EAGER iteration

ISPN-2619 Avoid empty checks in test using method specific keys

ISPN-2859 Reduce the verbosity of state transfer trace logs

ISPN-2848 Use the new bundling mechanism from JGroups 3.3.0

* set new bundler in jgroups configuration files

* removed deprecated flags

* remove unused parameters

    • -3
    • +2
    /core/src/main/resources/jgroups-ec2.xml
    • -3
    • +2
    /core/src/main/resources/jgroups-tcp.xml
    • -3
    • +2
    /core/src/main/resources/jgroups-udp.xml
    • -3
    • +2
    /core/src/test/resources/stacks/tcp.xml
    • -2
    • +2
    /core/src/test/resources/stacks/udp.xml
ISPN-2846 OperationsDuringStateTransferTest fails if run on non-tx cache

* Refactor OperationsDuringStateTransferTest into sublasses to test with and without TX; also add tests for put and remove operations

* NonTxDistributionInterceptor fetches remotely the value before executing a write command if the transfer of the key is in progress and the key is not available in the local data container. If the remote fetch returns null a local lookup is reattempted.

Various interceptor cleanups

* Remove unnecessary members in NonTransactionalLockingInterceptor

* Make some members private in TxDistributionInterceptor

* Fix typos in ConditionalOperationsConcurrentTest and make some members final

* Remove unused methods from BaseDistFunctionalTest and move method BaseDistFunctionalTest.createNewConsistentHash to ConsistentHashPerfTest where it is used

ISPN-2796 ConcurrentCacheManagerTest.testConcurrentGetCacheCalls hangs the test suite (randomly)

* DefaultCacheManager.wireAndStartCache didn't open the cache latch if

it failed to inject all dependencies, blocking all other threads.

* ComponentRegistry.getOrCreateComponent was not atomic if the component

was cache-scoped and the factory was global-scoped.

ISPN-2796 ConcurrentCacheManagerTest.testConcurrentGetCacheCalls hangs the test suite (randomly)

* DefaultCacheManager.wireAndStartCache didn't open the cache latch if

it failed to inject all dependencies, blocking all other threads.

* ComponentRegistry.getOrCreateComponent was not atomic if the component

was cache-scoped and the factory was global-scoped.

HHH-8010 Support app-managed EntityManagerFactory and SessionFactory creation by OSGi bundles

Conflicts:

hibernate-core/src/main/java/org/hibernate/boot/registry/classloading/internal/ClassLoaderServiceImpl.java

hibernate-core/src/main/java/org/hibernate/cfg/Configuration.java

hibernate-core/src/main/java/org/hibernate/engine/jdbc/connections/internal/DriverManagerConnectionProviderImpl.java

hibernate-entitymanager/src/main/java/org/hibernate/ejb/Ejb3Configuration.java

hibernate-osgi/src/main/java/org/hibernate/osgi/HibernateBundleActivator.java