Avoiding failure when using frozen indices (#1842) (#1886)

Author: masseyke

mr/src/main/java/org/elasticsearch/hadoop/rest/RestRepository.java

@@ -311,14 +311,16 @@ Scroll scroll(String query, BytesArray body, ScrollReader reader) throws IOExcep
         InputStream scroll = client.execute(POST, query, body).body();
         try {
             Scroll scrollResult = reader.read(scroll);
-            if (settings.getInternalVersionOrThrow().onOrBefore(EsMajorVersion.V_2_X)) {
+            if (scrollResult == null) {
+                log.info(String.format("No scroll for query [%s/%s], likely because the index is frozen", query, body));
+            } else if (settings.getInternalVersionOrThrow().onOrBefore(EsMajorVersion.V_2_X)) {
                 // On ES 2.X and before, a scroll response does not contain any hits to start with.
                 // Another request will be needed.
                 scrollResult = new Scroll(scrollResult.getScrollId(), scrollResult.getTotalHits(), false);
             }
             return scrollResult;
         } finally {
-            if (scroll instanceof StatsAware) {
+            if (scroll != null && scroll instanceof StatsAware) {
                 stats.aggregate(((StatsAware) scroll).stats());
             }
         }

mr/src/main/java/org/elasticsearch/hadoop/rest/ScrollQuery.java

@@ -92,6 +92,10 @@ public boolean hasNext() {
 
             try {
                 Scroll scroll = repository.scroll(query, body, reader);
+                if (scroll == null) {
+                    finished = true;
+                    return false;
+                }
                 // size is passed as a limit (since we can't pass it directly into the request) - if it's not specified (<1) just scroll the whole index
                 size = (size < 1 ? scroll.getTotalHits() : size);
                 scrollId = scroll.getScrollId();
@@ -115,6 +119,10 @@ public boolean hasNext() {
 
             try {
                 Scroll scroll = repository.scroll(scrollId, reader);
+                if (scroll == null) {
+                    finished = true;
+                    return false;
+                }
                 scrollId = scroll.getScrollId();
                 batch = scroll.getHits();
                 finished = scroll.isConcluded();

mr/src/main/java/org/elasticsearch/hadoop/serialization/ScrollReader.java

@@ -268,6 +268,12 @@ public Scroll read(InputStream content) throws IOException {
     private Scroll read(Parser parser, BytesArray input) {
         // get scroll_id
         Token token = ParsingUtils.seek(parser, SCROLL_ID);
+        if (token == null) { // no scroll id is returned for frozen indices
+            if (log.isTraceEnabled()) {
+                log.info("No scroll id found, likely because the index is frozen");
+            }
+            return null;
+        }
         Assert.isTrue(token == Token.VALUE_STRING, "invalid response");
         String scrollId = parser.text();
 

mr/src/test/java/org/elasticsearch/hadoop/rest/ScrollQueryTest.java

@@ -97,6 +97,35 @@ private RestRepository mockRepository(boolean firstScrollReturnsHits) throws Exc
 
 
 
+        return mocked;
+    }
+
+    @Test
+    public void testFrozen() throws Exception {
+        // Frozen indices return a null scroll
+        RestRepository repository = mockRepositoryFrozenIndex();
+        ScrollReader scrollReader = Mockito.mock(ScrollReader.class);
+
+        String query = "/index/type/_search?scroll=10m&etc=etc";
+        BytesArray body = new BytesArray("{}");
+        long size = 100;
+
+        ScrollQuery scrollQuery = new ScrollQuery(repository, query, body, size, scrollReader);
+
+        Assert.assertFalse(scrollQuery.hasNext());
+        scrollQuery.close();
+        Mockito.verify(repository).close();
+        Stats stats = scrollQuery.stats();
+        Assert.assertEquals(0, stats.docsReceived);
+    }
+
+    private RestRepository mockRepositoryFrozenIndex() throws Exception {
+        RestRepository mocked = Mockito.mock(RestRepository.class);
+        Mockito.doReturn(null).when(mocked).scroll(Matchers.anyString(), Matchers.any(BytesArray.class), Matchers.any(ScrollReader.class));
+        RestClient mockClient = Mockito.mock(RestClient.class);
+        Mockito.when(mockClient.deleteScroll(Matchers.eq("mnop"))).thenReturn(true);
+        Mockito.when(mockClient.deleteScroll(Matchers.anyString())).thenReturn(false);
+        Mockito.doReturn(mockClient).when(mocked).getRestClient();
         return mocked;
     }
 }

mr/src/test/java/org/elasticsearch/hadoop/serialization/ScrollReaderTest.java

@@ -585,6 +585,22 @@ public void testScrollWithHandlersThatCorrectsError() throws IOException {
         assertEquals(4L, JsonUtils.query("number").apply(scroll.getHits().get(0)[1]));
     }
 
+    @Test
+    public void testNoScrollIdFromFrozenIndex() throws IOException {
+        MappingSet mappings = getMappingSet("numbers-as-strings"); // The schema doesn't matter since there's no data
+        InputStream stream = getClass().getResourceAsStream(scrollData("no-scroll-id"));
+        Settings testSettings = new TestSettings();
+        testSettings.setProperty(ConfigurationOptions.ES_READ_METADATA, "" + readMetadata);
+        testSettings.setProperty(ConfigurationOptions.ES_READ_METADATA_FIELD, "" + metadataField);
+        testSettings.setProperty(ConfigurationOptions.ES_OUTPUT_JSON, "" + readAsJson);
+        testSettings.setProperty(DeserializationHandlerLoader.ES_READ_DATA_ERROR_HANDLERS , "fix");
+        testSettings.setProperty(DeserializationHandlerLoader.ES_READ_DATA_ERROR_HANDLER + ".fix" , CorrectingHandler.class.getName());
+        JdkValueReader valueReader = ObjectUtils.instantiate(JdkValueReader.class.getName(), testSettings);
+        ScrollReader reader = new ScrollReader(ScrollReaderConfigBuilder.builder(valueReader, mappings.getResolvedView(), testSettings));
+        ScrollReader.Scroll scroll = reader.read(stream);
+        assertNull(scroll);
+    }
+
     /**
      * Case: Handler throws random Exceptions
      * Outcome: Processing fails fast.

mr/src/test/resources/org/elasticsearch/hadoop/serialization/scrollReaderTestData/no-scroll-id/scroll.json

@@ -0,0 +1,20 @@
+{
+  "took":0,
+  "timed_out":false,
+  "_shards":{
+    "total":0,
+    "successful":0,
+    "skipped":0,
+    "failed":0
+  },
+  "hits":{
+    "total":{
+      "value":0,
+      "relation":"eq"
+    },
+    "max_score":0.0,
+    "hits":[
+
+    ]
+  }
+}