From 74634eb002717a9684d00f0e9dc263ab7eb49246 Mon Sep 17 00:00:00 2001
From: Wei-Chiu Chuang <weichiu@apache.org>
Date: Fri, 23 Oct 2020 12:05:53 -0700
Subject: [PATCH] HDFS-15644. Failed volumes can cause DNs to stop block
 reporting. Contributed by Ahmed Hussein.

---
 .../fsdataset/impl/FsDatasetImpl.java         | 38 ++++++++++---------
 1 file changed, 21 insertions(+), 17 deletions(-)

diff --git a/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/fsdataset/impl/FsDatasetImpl.java b/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/fsdataset/impl/FsDatasetImpl.java
index e65e37fe13d..ed779923235 100644
--- a/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/fsdataset/impl/FsDatasetImpl.java
+++ b/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/fsdataset/impl/FsDatasetImpl.java
@@ -1957,28 +1957,32 @@ class FsDatasetImpl implements FsDatasetSpi<FsVolumeImpl> {
           continue;
         }
         String volStorageID = b.getVolume().getStorageID();
-        if (!builders.containsKey(volStorageID)) {
+        switch(b.getState()) {
+        case FINALIZED:
+        case RBW:
+        case RWR:
+          break;
+        case RUR:
+          // use the original replica.
+          b = b.getOriginalReplica();
+          break;
+        case TEMPORARY:
+          continue;
+        default:
+          assert false : "Illegal ReplicaInfo state.";
+          continue;
+        }
+        BlockListAsLongs.Builder storageBuilder = builders.get(volStorageID);
+        // a storage in the process of failing will not be in the volumes list
+        // but will be in the replica map.
+        if (storageBuilder != null) {
+          storageBuilder.add(b);
+        } else {
           if (!missingVolumesReported.contains(volStorageID)) {
             LOG.warn("Storage volume: " + volStorageID + " missing for the"
                 + " replica block: " + b + ". Probably being removed!");
             missingVolumesReported.add(volStorageID);
           }
-          continue;
-        }
-        switch(b.getState()) {
-        case FINALIZED:
-        case RBW:
-        case RWR:
-          builders.get(volStorageID).add(b);
-          break;
-        case RUR:
-          ReplicaInfo orig = b.getOriginalReplica();
-          builders.get(volStorageID).add(orig);
-          break;
-        case TEMPORARY:
-          break;
-        default:
-          assert false : "Illegal ReplicaInfo state.";
         }
       }
     }