big-data/hadoop
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

HDFS-16540 Data locality is lost when DataNode pod restarts in kubernetes (#4170)

Browse Source 
When DN with the same UUID is registered with a different IP, host2DatanodeMap needs to be updated accordingly.
This commit is contained in:
huaxiangsun 2022-04-28 10:24:37 -07:00 committed by GitHub
parent 63187083cc
commit bda0881bf9
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 45 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

BIN
.BUILDING.txt.swp Normal file
View File

Binary file not shown.

17
hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/blockmanagement/DatanodeManager.java
View File

@ -1171,6 +1171,7 @@ public void registerDatanode(DatanodeRegistration nodeReg)
nodeN = null;
}
boolean updateHost2DatanodeMap = false;
if (nodeS != null) {
if (nodeN == nodeS) {
// The same datanode has been just restarted to serve the same data
@ -1189,7 +1190,11 @@ public void registerDatanode(DatanodeRegistration nodeReg)
nodes with its data cleared (or user can just remove the StorageID
value in "VERSION" file under the data directory of the datanode,
but this is might not work if VERSION file format has changed
*/
*/
// Check if nodeS's host information is same as nodeReg's, if not,
// it needs to update host2DatanodeMap accordringly.
updateHost2DatanodeMap = !nodeS.getXferAddr().equals(nodeReg.getXferAddr());
NameNode.stateChangeLog.info("BLOCK* registerDatanode: " + nodeS
+ " is replaced by " + nodeReg + " with the same storageID "
+ nodeReg.getDatanodeUuid());
@ -1199,6 +1204,11 @@ nodes with its data cleared (or user can just remove the StorageID
try {
// update cluster map
getNetworkTopology().remove(nodeS);
// Update Host2DatanodeMap
if (updateHost2DatanodeMap) {
getHost2DatanodeMap().remove(nodeS);
}
if(shouldCountVersion(nodeS)) {
decrementVersionCount(nodeS.getSoftwareVersion());
}
@ -1217,6 +1227,11 @@ nodes with its data cleared (or user can just remove the StorageID
nodeS.setDependentHostNames(
getNetworkDependenciesWithDefault(nodeS));
}
if (updateHost2DatanodeMap) {
getHost2DatanodeMap().add(nodeS);
}
getNetworkTopology().add(nodeS);
resolveUpgradeDomain(nodeS);

29
hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/blockmanagement/TestDatanodeManager.java
View File

@ -138,6 +138,35 @@ public void testNumVersionsCorrectAfterReregister()
mapToCheck.get("version1").intValue(), 1);
}
/**
* This test checks that if a node is re-registered with a different ip, its
* host2DatanodeMap is correctly updated with the new ip.
*/
@Test
public void testHost2NodeMapCorrectAfterReregister()
throws IOException, InterruptedException {
//Create the DatanodeManager which will be tested
FSNamesystem fsn = Mockito.mock(FSNamesystem.class);
Mockito.when(fsn.hasWriteLock()).thenReturn(true);
Configuration conf = new Configuration();
DatanodeManager dm = mockDatanodeManager(fsn, conf);
String storageID = "someStorageID1";
String ipOld = "someIPOld" + storageID;
String ipNew = "someIPNew" + storageID;
dm.registerDatanode(new DatanodeRegistration(
new DatanodeID(ipOld, "", storageID, 9000, 0, 0, 0),
null, null, "version"));
dm.registerDatanode(new DatanodeRegistration(
new DatanodeID(ipNew, "", storageID, 9000, 0, 0, 0),
null, null, "version"));
assertNull("should be no node with old ip", dm.getDatanodeByHost(ipOld));
assertNotNull("should be a node with new ip", dm.getDatanodeByHost(ipNew));
}
/**
* This test sends a random sequence of node registrations and node removals
* to the DatanodeManager (of nodes with different IDs and versions), and