[
https://issues.apache.org/jira/browse/HDFS-17093?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17748370#comment-17748370
]
ASF GitHub Bot commented on HDFS-17093:
---------------------------------------
zhangshuyan0 commented on code in PR #5855:
URL: https://github.com/apache/hadoop/pull/5855#discussion_r1277008543
##########
hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/blockmanagement/TestBlockReportLease.java:
##########
@@ -269,4 +272,88 @@ private StorageBlockReport[]
createReports(DatanodeStorage[] dnStorages,
}
return storageBlockReports;
}
+
+ @Test
+ public void testFirstIncompleteBlockReport() throws Exception {
+ HdfsConfiguration conf = new HdfsConfiguration();
+ Random rand = new Random();
+
+ try (MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf)
+ .numDataNodes(1).build()) {
+ cluster.waitActive();
+
+ FSNamesystem fsn = cluster.getNamesystem();
+
+ NameNode nameNode = cluster.getNameNode();
+ // pretend to be in safemode
+ NameNodeAdapter.enterSafeMode(nameNode, false);
+
+ BlockManager blockManager = fsn.getBlockManager();
+ BlockManager spyBlockManager = spy(blockManager);
+ fsn.setBlockManagerForTesting(spyBlockManager);
+ String poolId = cluster.getNamesystem().getBlockPoolId();
+
+ NamenodeProtocols rpcServer = cluster.getNameNodeRpc();
+
+ // Test based on one DataNode report to Namenode
+ DataNode dn = cluster.getDataNodes().get(0);
+ DatanodeDescriptor datanodeDescriptor = spyBlockManager
+ .getDatanodeManager().getDatanode(dn.getDatanodeId());
+
+ DatanodeRegistration dnRegistration = dn.getDNRegistrationForBP(poolId);
+ StorageReport[] storages = dn.getFSDataset().getStorageReports(poolId);
+
+ // Send heartbeat and request full block report lease
+ HeartbeatResponse hbResponse = rpcServer.sendHeartbeat(
+ dnRegistration, storages, 0, 0, 0, 0, 0, null, true,
+ SlowPeerReports.EMPTY_REPORT, SlowDiskReports.EMPTY_REPORT);
+
+ DelayAnswer delayer = new DelayAnswer(BlockManager.LOG);
+ doAnswer(delayer).when(spyBlockManager).processReport(
+ any(DatanodeStorageInfo.class),
+ any(BlockListAsLongs.class));
+
+ ExecutorService pool = Executors.newFixedThreadPool(1);
+
+ // Trigger sendBlockReport
+ BlockReportContext brContext = new BlockReportContext(1, 0,
+ rand.nextLong(), hbResponse.getFullBlockReportLeaseId());
+ // Build every storage with 100 blocks for sending report
+ DatanodeStorage[] datanodeStorages
+ = new DatanodeStorage[storages.length];
+ for (int i = 0; i < storages.length; i++) {
+ datanodeStorages[i] = storages[i].getStorage();
+ StorageBlockReport[] reports = createReports(datanodeStorages, 100);
+
+ // The first multiple send once, simulating the failure of the first
report, only send successfully once
+ if(i == 0){
+ rpcServer.blockReport(dnRegistration, poolId, reports, brContext);
+ }
+
+ // Send blockReport
+ DatanodeCommand datanodeCommand =
rpcServer.blockReport(dnRegistration, poolId, reports,
+ brContext);
+
+ // Wait until BlockManager calls processReport
+ delayer.waitForCall();
+
+ // Remove full block report lease about dn
+ spyBlockManager.getBlockReportLeaseManager()
+ .removeLease(datanodeDescriptor);
Review Comment:
The problem in this UT is the same as before, you still actively call
`removeLease` in the code, which doesn't seem to happen in real code. It
remains confusing why `removeLease` is called when the first block report was
not successfully processed.
> In the case of all datanodes sending FBR when the namenode restarts (large
> clusters), there is an issue with incomplete block reporting
> ---------------------------------------------------------------------------------------------------------------------------------------
>
> Key: HDFS-17093
> URL: https://issues.apache.org/jira/browse/HDFS-17093
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: namenode
> Affects Versions: 3.3.4
> Reporter: Yanlei Yu
> Priority: Minor
> Labels: pull-request-available
>
> In our cluster of 800+ nodes, after restarting the namenode, we found that
> some datanodes did not report enough blocks, causing the namenode to stay in
> secure mode for a long time after restarting because of incomplete block
> reporting
> I found in the logs of the datanode with incomplete block reporting that the
> first FBR attempt failed, possibly due to namenode stress, and then a second
> FBR attempt was made as follows:
> {code:java}
> ....
> 2023-07-17 11:29:28,982 INFO org.apache.hadoop.hdfs.server.datanode.DataNode:
> Unsuccessfully sent block report 0x6237a52c1e817e, containing 12 storage
> report(s), of which we sent 1. The reports had 1099057 total blocks and used
> 1 RPC(s). This took 294 msec to generate and 101721 msecs for RPC and NN
> processing. Got back no commands.
> 2023-07-17 11:37:04,014 INFO org.apache.hadoop.hdfs.server.datanode.DataNode:
> Successfully sent block report 0x62382416f3f055, containing 12 storage
> report(s), of which we sent 12. The reports had 1099048 total blocks and used
> 12 RPC(s). This took 295 msec to generate and 11647 msecs for RPC and NN
> processing. Got back no commands. {code}
> There's nothing wrong with that. Retry the send if it fails But on the
> namenode side of the logic:
> {code:java}
> if (namesystem.isInStartupSafeMode()
> && !StorageType.PROVIDED.equals(storageInfo.getStorageType())
> && storageInfo.getBlockReportCount() > 0) {
> blockLog.info("BLOCK* processReport 0x{} with lease ID 0x{}: "
> + "discarded non-initial block report from {}"
> + " because namenode still in startup phase",
> strBlockReportId, fullBrLeaseId, nodeID);
> blockReportLeaseManager.removeLease(node);
> return !node.hasStaleStorages();
> } {code}
> When a disk was identified as the report is not the first time, namely
> storageInfo. GetBlockReportCount > 0, Will remove the ticket from the
> datanode, lead to a second report failed because no lease
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
