Class RollingBatchRestartRsAction
java.lang.Object
org.apache.hadoop.hbase.chaos.actions.Action
org.apache.hadoop.hbase.chaos.actions.RestartActionBaseAction
org.apache.hadoop.hbase.chaos.actions.BatchRestartRsAction
org.apache.hadoop.hbase.chaos.actions.RollingBatchRestartRsAction
Restarts a ratio of the regionservers in a rolling fashion. At each step, either kills a server,
or starts one, sleeping randomly (0-sleepTime) in between steps. The parameter maxDeadServers
limits the maximum number of servers that can be down at the same time during rolling restarts.
-
Nested Class Summary
Nested ClassesModifier and TypeClassDescription(package private) static enumNested classes/interfaces inherited from class org.apache.hadoop.hbase.chaos.actions.Action
Action.ActionContext -
Field Summary
FieldsModifier and TypeFieldDescriptionprivate static final org.slf4j.Loggerprotected intFields inherited from class org.apache.hadoop.hbase.chaos.actions.BatchRestartRsAction
ratioFields inherited from class org.apache.hadoop.hbase.chaos.actions.RestartActionBaseAction
sleepTimeFields inherited from class org.apache.hadoop.hbase.chaos.actions.Action
cluster, context, initialServers, initialStatus, KILL_DATANODE_TIMEOUT_DEFAULT, KILL_DATANODE_TIMEOUT_KEY, KILL_MASTER_TIMEOUT_DEFAULT, KILL_MASTER_TIMEOUT_KEY, KILL_NAMENODE_TIMEOUT_DEFAULT, KILL_NAMENODE_TIMEOUT_KEY, KILL_RS_TIMEOUT_DEFAULT, KILL_RS_TIMEOUT_KEY, KILL_ZK_NODE_TIMEOUT_DEFAULT, KILL_ZK_NODE_TIMEOUT_KEY, killDataNodeTimeout, killMasterTimeout, killNameNodeTimeout, killRsTimeout, killZkNodeTimeout, monkeyProps, skipMetaRS, START_DATANODE_TIMEOUT_DEFAULT, START_DATANODE_TIMEOUT_KEY, START_MASTER_TIMEOUT_DEFAULT, START_MASTER_TIMEOUT_KEY, START_NAMENODE_TIMEOUT_DEFAULT, START_NAMENODE_TIMEOUT_KEY, START_RS_TIMEOUT_DEFAULT, START_RS_TIMEOUT_KEY, START_ZK_NODE_TIMEOUT_DEFAULT, START_ZK_NODE_TIMEOUT_KEY, startDataNodeTimeout, startMasterTimeout, startNameNodeTimeout, startRsTimeout, startZkNodeTimeout -
Constructor Summary
ConstructorsConstructorDescriptionRollingBatchRestartRsAction(long sleepTime, float ratio) RollingBatchRestartRsAction(long sleepTime, float ratio, int maxDeadServers) RollingBatchRestartRsAction(long sleepTime, float ratio, int maxDeadServers, boolean skipMetaRS) -
Method Summary
Methods inherited from class org.apache.hadoop.hbase.chaos.actions.RestartActionBaseAction
gracefulRestartRs, restartDataNode, restartJournalNode, restartMaster, restartNameNode, restartRs, restartZKNode, sleepMethods inherited from class org.apache.hadoop.hbase.chaos.actions.Action
forceBalancer, getConf, getCurrentServers, init, killDataNode, killJournalNode, killMaster, killNameNode, killRs, killZKNode, modifyAllTableColumns, modifyAllTableColumns, resumeRs, setBalancer, startDataNode, startJournalNode, startMaster, startNameNode, startRs, startZKNode, stopRs, suspendRs, unbalanceRegions
-
Field Details
-
LOG
-
maxDeadServers
-
-
Constructor Details
-
RollingBatchRestartRsAction
-
RollingBatchRestartRsAction
-
RollingBatchRestartRsAction
public RollingBatchRestartRsAction(long sleepTime, float ratio, int maxDeadServers, boolean skipMetaRS)
-
-
Method Details
-
getLogger
Description copied from class:ActionRetrieve the instance'sLogger, for use throughout the class hierarchy.- Overrides:
getLoggerin classBatchRestartRsAction
-
perform
- Overrides:
performin classBatchRestartRsAction- Throws:
Exception
-
selectServers
- Throws:
IOException
-
main
Small test to ensure the class basically works.- Throws:
Exception
-