----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://review.cloudera.org/r/484/ -----------------------------------------------------------
(Updated 2010-08-05 00:19:37.074497) Review request for hbase, stack and Karthik Ranganathan. Changes ------- notes from 8/4 (with what i did tonight for them, which is most of what is different in this update) --- * in CatalogTracker need to stabilize on one getRoot and one getMeta method to use that waits and uses the default wait-for-catalogs timeout. We should get rid of the 'refresh' boolean that I have in there and should always ping the server to ensure it is serving the region before we return it. If we do eventually drop root and put the meta locations into zk we would no longer need this, so will not always have to pay this tax. >> This is done. You pass default timeout in constructor. Two methods now are: waitForRootServerConnectionDefault() waitForMetaServerConnectionDefault() * ROOT changes RootEditor -> RootLocationEditor, delete -> unset Change the way we unset the root location. Set the data to null rather than deleting the node. Requires changes to RootLocationEditor and RootRegionTracker. >> Thought there was a race condition here, but there is not. In fact, we do not even need to set the watch in the delete method. It is already properly being handled by RootRegionTracker. * In AssignmentManager.processFailure() need to insert RegionState into RIT map >> This is done. This needs tests but I think failover is all in place now. * On RS-side, make separate OpenRootHandler and OpenMetaHandler >> Added four new handlers for open/close of root/meta and associated executors * Add priorities to Opened/Closed handlers on Master >> Added ROOT, META, USER priorities * In RegionTransitionData, store the actual byte[] regionName rather than the encoded name >> Done. We should also get in practice of naming variables encodedName if it is that. * Executor services need to be using a priority queue >> Done. I think all stuff to set pool size and add priorities is in. * In EventType, completely remove differentiating between Master and RS. This means for a given EventType, it will map to the same handler whether it is on RS or Master. >> Done. Also in EventType, remove fromByte() and use an ordinal() method >> Done. Can we remove even having the (int) values for the enums now? Summary ------- This is basically the master rewrite. It's massive, I apologize. I can start a cluster, create a table, insert a row, and get the row. Restarting should work too. I completely ripped out BaseScanner/MetaScanner as well as the RegionServerOperationQueue stuff. I'm sold on the former but there's a bit of work to fully replace the latter though I think it's a desirable direction to get rid of it. The RSOQ stuff has been replaced with EventHandlers. The BaseScanner is no longer necessary because we are not utilizing META to determine when things are unassigned. All the table operations (enable/disable, create/delete, addcolumn/del/modify) have been reimplemented in handlers. All META/ROOT server access is done through the new CatalogManager. This uses ZK and ROOT/META to keep active track of the current locations of the catalog regions. A bit more work needs to be done on defining our retry semantics for different operations. The actual read/write operations on ROOT/META are done with new classes RootEditor, MetaEditor, and MetaReader. A lot of this code was scattered around HMaster, RegionManager, ServerManager, HRegion, etc... Some new methods in FileSystemManager, collected from being scattered around. This also does final straw to old ZKWrapper and removes it. All ZK stuff is going through the new stuff. Open/close is now going via RPC and ZK, no longer piggybacking any of these messages on heartbeats (and with splits not working right now, even that is gone, so there will be nothing it seems) RegionManager was completely removed and replaced with a shiny, brand new AssignmentManager. It's already managed to get a bit hectic but is pretty straightforward. It is the heart of zk-based region assignment and most logic is in here. It contains regionsInTransition, the region plans, etc... I added a new /disabled node in ZK to keep track of tables that are disabled. I think we need this to be able to reliably do disables across failures. Among other changes.... Here's my rough list of things left to do (covers most but not all TODOs in the branch) * make final decisions on root/meta timeouts. almost everyone is coordinating access through CatalogTracker which should make it easy to standardize. if there are operations that should just retry indefinitely, they need to resubmit themselves to their executor service. * move splits to RS side, integrate new patch from stack on trunk might need a new CREATED unassigned now, or new rpc, but get rid of sending split notification on heartbeat? how to handle splits concurrent with disable? * review master startup order * figure what to do with client table admin ops (flush, split, compact) * MasterAddressManager move to extending ZooKeeperNodeTracker * on region open (and wherever split children notify master) should check if if the table is disabled and should close the regions... maybe. * regionserver exit needs to be reimplemented in servermanager also rs expiration * add priorities and pool size to handlers (bring in from flush patch) * in RootEditor there is a race condition between delete and watch? * migrate TestZKBased* tests to use new handlers * make sync calls for enable/disable (check and verify methods?) * integrate load balancing * finish TODOs on new failover path and remove old code in joinExistingCluster() * finish TODOs on timeout monitor in assignmentmanager * review filesystemmanager calls * figure how to handle the very rare but possible race condition where two RSs will update META and the later one can squash the valid one if there was a long gc pause * synchronize all access to the boolean in ActiveMasterManager (now this is probably just move it to extend ZKNodeTracker) * there are some races with master wanting to connect for rpc to regionserver and the rs starting its rpc server, need to address * migrate TestMasterTransitions or make new? * fix or remove last couple master tests that used RSOQ * write new tests!!! This addresses bug HBASE-2697. http://issues.apache.org/jira/browse/HBASE-2697 Diffs (updated) ----- branches/0.90_master_rewrite/BRANCH_TODO.txt 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/Abortable.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/ClusterStatus.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/HRegionInfo.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/HServerInfo.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/InvalidFamilyOperationException.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/ServerController.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/catalog/CatalogTracker.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/catalog/MetaEditor.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/catalog/MetaReader.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/catalog/RootLocationEditor.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/client/HBaseAdmin.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/client/HConnection.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/client/HConnectionManager.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/client/MetaScanner.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/executor/EventHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/executor/HBaseEventHandler.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/executor/HBaseExecutorService.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/executor/RegionTransitionData.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/ipc/HMasterInterface.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/ipc/HRegionInterface.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/ActiveMasterManager.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/AddColumn.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/AssignmentManager.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/BaseScanner.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/ChangeTableState.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/ColumnOperation.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/DeleteColumn.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/FileSystemManager.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/HMaster.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/LoadBalancer.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/MasterController.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/MasterStatus.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/MetaRegion.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/MetaScanner.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/ModifyColumn.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/ModifyTableMeta.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/ProcessRegionClose.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/ProcessRegionOpen.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/ProcessRegionStatusChange.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/ProcessServerShutdown.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/RegionManager.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/RegionServerOperation.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/RegionServerOperationListener.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/RegionServerOperationQueue.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/RetryableMetaOperation.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/RootScanner.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/ServerManager.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/TableDelete.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/TableOperation.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/ClosedRegionHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/DeleteTableHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/DisableTableHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/EnableTableHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/MasterCloseRegionHandler.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/MasterOpenRegionHandler.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/ModifyTableHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/OpenedRegionHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/ServerShutdownHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/TableAddFamilyHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/TableDeleteFamilyHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/TableEventHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/master/handler/TableModifyFamilyHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/CompactSplitThread.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/HRegion.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/HRegionServer.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/MasterAddressManager.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/RSZookeeperUpdater.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/RegionServerController.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/handler/CloseMetaHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/handler/CloseRegionHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/handler/CloseRootHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/handler/OpenMetaHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/handler/OpenRegionHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/regionserver/handler/OpenRootHandler.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/zookeeper/ClusterStatusTracker.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/zookeeper/MetaNodeTracker.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/zookeeper/RootRegionTracker.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/zookeeper/ZKAssign.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/zookeeper/ZKTableDisable.java PRE-CREATION branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/zookeeper/ZKUtil.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/zookeeper/ZooKeeperListener.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/zookeeper/ZooKeeperNodeTracker.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/zookeeper/ZooKeeperWatcher.java 979909 branches/0.90_master_rewrite/src/main/java/org/apache/hadoop/hbase/zookeeper/ZooKeeperWrapper.java 979909 branches/0.90_master_rewrite/src/test/java/org/apache/hadoop/hbase/HBaseTestingUtility.java 979909 branches/0.90_master_rewrite/src/test/java/org/apache/hadoop/hbase/TestZooKeeper.java 979909 branches/0.90_master_rewrite/src/test/java/org/apache/hadoop/hbase/master/TestActiveMasterManager.java 979909 branches/0.90_master_rewrite/src/test/java/org/apache/hadoop/hbase/master/TestMaster.java 979909 branches/0.90_master_rewrite/src/test/java/org/apache/hadoop/hbase/master/TestRestartCluster.java 979909 branches/0.90_master_rewrite/src/test/java/org/apache/hadoop/hbase/master/TestZKBasedCloseRegion.java 979909 branches/0.90_master_rewrite/src/test/java/org/apache/hadoop/hbase/master/TestZKBasedReopenRegion.java 979909 branches/0.90_master_rewrite/src/test/java/org/apache/hadoop/hbase/regionserver/TestMasterAddressManager.java 979909 branches/0.90_master_rewrite/src/test/java/org/apache/hadoop/hbase/util/TestBytes.java 979909 Diff: http://review.cloudera.org/r/484/diff Testing ------- simple cluster tests passing, need to run (and write) more tests Thanks, Jonathan