Package org.apache.hadoop.hbase.master
package org.apache.hadoop.hbase.master
-
ClassDescriptionBase class for testing distributed log splitting.An implementation of HMaster that always runs as a stand by and never transitions to active.An implementation of ActiveMasterManager that never transitions it's master to active state.A mock RegionServer implementation.Test the
ActiveMasterManager.Dummy Master Implementation.Test balancer with disabled tableTrivial test to confirm that we do not get 0 infoPort.Trivial test to confirm that we can get last flushed sequence id by encodedRegionName.Testcase for HBASE-21490.Define this class because the test needs to overrideStochasticLoadBalancer.balanceTable(org.apache.hadoop.hbase.TableName, java.util.Map<org.apache.hadoop.hbase.ServerName, java.util.List<org.apache.hadoop.hbase.client.RegionInfo>>), which is protected.Tests to validate if HMaster default chores are scheduledTestMasterChoreScheduled.TestChoreField<E extends org.apache.hadoop.hbase.ScheduledChore>Reflect into theHMasterinstance and find by field name a specified instance ofScheduledChore.Tests that the MasterRpcServices is correctly searching for implementations of the Coprocessor Service and not just the "default" implementations of those services.Test the master filesystem in a local clusterTest the master filesystem in a local cluster with Store File Tracking explicitly set in global configTest the master filesystem in a local clusterMake open region very slowStandup the master and fake it to test various aspects of master function.Test transitions of state across the master.Tests handling of meta-carrying region server failover.Test for MetricsMasterProcSourceImplTest for the CompatibilitySingletonFactory and building MetricsMasterSourceTest for MetricsMasterSourceImplTestcase for HBASE-26193.Tests that table state is mirrored out to zookeeper for hbase-1.x clients.Tests for OldWALsDirSizeChore Here we are using theMockMasterServicesto mock the Hbase Master.Test reuse storefiles within data directory when cluster failover with a set of new region servers with different hostnames with or without WALs and Zookeeper ZNodes, the master and cluster should fail respectively if there is any situation considered as not supported.Test for RegionsRecoveryChoreSimple helper class that just keeps track of whether or not its stopped.Test for Regions Recovery Config ManagerSimulate the scenario described in HBASE-26245, where we clean the WAL directory and try to start the cluster.Tests the restarting of everything as done during rolling restarts.Testcase for HBASE-20634Test to confirm that we will not hang when stop a backup master which is trying to become the active master.Testcase to confirm that we will not hang when shutdown a cluster with no live region servers.Tests the default table lock managerRun tests that use the HBase clients;TableBuilder.