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

HDFS-15095. Fix TestDecommissioningStatus. Contributed by Ahmed Hussein.

Browse Source
This commit is contained in:
Kihwal Lee 2020-01-10 12:43:21 -06:00
parent c4fb43c942
commit 5fb901ac40
1 changed files with 30 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/namenode/TestDecommissioningStatus.java
View File

@ -27,6 +27,7 @@
import java.util.Arrays;
import java.util.List;
import java.util.concurrent.TimeoutException;
import org.apache.commons.io.output.ByteArrayOutputStream;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.BlockLocation;
@ -55,6 +56,7 @@
import org.apache.hadoop.hdfs.server.datanode.DataNodeTestUtils;
import org.apache.hadoop.hdfs.tools.DFSAdmin;
import org.apache.hadoop.hdfs.util.HostsFileWriter;
import org.apache.hadoop.test.GenericTestUtils;
import org.apache.log4j.Level;
import org.apache.log4j.Logger;
import org.junit.After;
@ -219,6 +221,23 @@ protected void checkDFSAdminDecommissionStatus(
}
}
/**
* Allows the main thread to block until the decommission is checked by the
* admin manager.
* @param dnAdminMgr admin instance in the datanode manager.
* @param trackedNumber number of nodes expected to be DECOMMISSIONED or
* IN_MAINTENANCE.
* @throws TimeoutException
* @throws InterruptedException
*/
private void waitForDecommissionedNodes(final DatanodeAdminManager dnAdminMgr,
final int trackedNumber)
throws TimeoutException, InterruptedException {
GenericTestUtils
.waitFor(() -> dnAdminMgr.getNumTrackedNodes() == trackedNumber,
100, 2000);
}
/**
* Tests Decommissioning Status in DFS.
*/
@ -254,11 +273,13 @@ public void testDecommissionStatus() throws Exception {
dm.refreshNodes(conf);
decommissionedNodes.add(downnode);
BlockManagerTestUtil.recheckDecommissionState(dm);
// Block until the admin's monitor updates the number of tracked nodes.
waitForDecommissionedNodes(dm.getDatanodeAdminManager(), iteration + 1);
final List<DatanodeDescriptor> decommissioningNodes = dm.getDecommissioningNodes();
if (iteration == 0) {
assertEquals(decommissioningNodes.size(), 1);
DatanodeDescriptor decommNode = decommissioningNodes.get(0);
// checkDecommissionStatus(decommNode, 3, 0, 1);
checkDecommissionStatus(decommNode, 3, 0, 1);
checkDFSAdminDecommissionStatus(decommissioningNodes.subList(0, 1),
fileSys, admin);
} else {
@ -330,11 +351,11 @@ public void testDecommissionStatusAfterDNRestart() throws Exception {
// Force DatanodeManager to check decommission state.
BlockManagerTestUtil.recheckDecommissionState(dm);
// Block until the admin's monitor updates the number of tracked nodes.
waitForDecommissionedNodes(dm.getDatanodeAdminManager(), 1);
// Verify that the DN remains in DECOMMISSION_INPROGRESS state.
assertTrue("the node should be DECOMMISSION_IN_PROGRESSS",
dead.get(0).isDecommissionInProgress());
// Check DatanodeManager#getDecommissionNodes, make sure it returns
// the node as decommissioning, even if it's dead
List<DatanodeDescriptor> decomlist = dm.getDecommissioningNodes();
@ -344,6 +365,8 @@ public void testDecommissionStatusAfterDNRestart() throws Exception {
// DECOMMISSION_IN_PROGRESS node become DECOMMISSIONED
AdminStatesBaseTest.cleanupFile(fileSys, f);
BlockManagerTestUtil.recheckDecommissionState(dm);
// Block until the admin's monitor updates the number of tracked nodes.
waitForDecommissionedNodes(dm.getDatanodeAdminManager(), 0);
assertTrue("the node should be decommissioned",
dead.get(0).isDecommissioned());
@ -378,6 +401,8 @@ public void testDecommissionDeadDN() throws Exception {
decommissionNode(dnName);
dm.refreshNodes(conf);
BlockManagerTestUtil.recheckDecommissionState(dm);
// Block until the admin's monitor updates the number of tracked nodes.
waitForDecommissionedNodes(dm.getDatanodeAdminManager(), 0);
assertTrue(dnDescriptor.isDecommissioned());
// Add the node back
@ -426,6 +451,8 @@ public void testDecommissionLosingData() throws Exception {
hostsFileWriter.initExcludeHosts(nodes);
dm.refreshNodes(conf);
BlockManagerTestUtil.recheckDecommissionState(dm);
// Block until the admin's monitor updates the number of tracked nodes.
waitForDecommissionedNodes(dm.getDatanodeAdminManager(), 0);
assertTrue(dnDescriptor0.isDecommissioned());
assertTrue(dnDescriptor1.isDecommissioned());