Do not send version conflicts to failure store (elastic#112537)

When indexing to a data stream with a failure store it's possible to get
a version conflict. The reproduction path is the following:

```
PUT /_bulk
{"create":{"_index": "my-ds-with-fs", "_id": "1"}}
{"@timestamp": "2022-01-01", "baz": "quick", "a": "brown", "b": "fox"}
{"create":{"_index": "my-ds-with-fs", "_id": "1"}}
{"@timestamp": "2022-01-01", "baz": "lazy", "a": "dog"}
```

We would like the second document to not be sent to the failure store
and return an error to the user:

```
{
  "errors" : true,
  "took" : 409,
  "items" : [
    {
      "create" : {
        "_index" : ".ds-my-ds-with-fs-xxxxx-xxxx",
        "_id" : "1",
        "_version" : 1,
        "result" : "created",
        "_shards" : {
          "total" : 2,
          "successful" : 1,
          "failed" : 0
        },
        "_seq_no" : 0,
        "_primary_term" : 1,
        "status" : 201
      }
    },
    {
      "create" : {
        "_index" : ".ds-my-ds-with-fs-xxxxx-xxxx",
        "_id" : "1",
        "status" : 409,
        "error" : {
          "type" : "version_conflict_engine_exception",
          "reason" : "[1]: version conflict, document already exists (current version [1])",
          "index_uuid" : ".....",
          "shard" : "0",
          "index" : ".ds-my-ds-with-fs-xxxxx-xxxx"
        }
      }
    }
  ]
}
```

The version conflict doc is counted as a rejected doc in APM telemetry.

modules/data-streams/src/yamlRestTest/resources/rest-api-spec/test/data_stream/190_failure_store_redirection.yml

@@ -748,3 +748,43 @@ teardown:
       indices.delete:
         index: .fs-logs-foobar-*
   - is_true: acknowledged
+
+---
+"Version conflicts are not redirected to failure store":
+  - requires:
+      cluster_features: ["gte_v8.16.0"]
+      reason: "Redirecting version conflicts to the failure store is considered a bug fixed in 8.16"
+      test_runner_features: [allowed_warnings, contains]
+
+  - do:
+      allowed_warnings:
+        - "index template [generic_logs_template] has index patterns [logs-*] matching patterns from existing older templates [global] with patterns (global => [*]); this template [generic_logs_template] will take precedence during new index creation"
+      indices.put_index_template:
+        name: generic_logs_template
+        body:
+          index_patterns: logs-*
+          data_stream:
+            failure_store: true
+          template:
+            settings:
+              number_of_shards:   1
+              number_of_replicas: 1
+            mappings:
+              properties:
+                '@timestamp':
+                  type: date
+                count:
+                  type: long
+
+  - do:
+      bulk:
+        refresh: true
+        body:
+          - '{ "create": { "_index": "logs-foobar", "_id": "1" } }'
+          - '{ "@timestamp": "2022-01-01", "baz": "quick", "a": "brown", "b": "fox" }'
+          - '{ "create": { "_index": "logs-foobar", "_id": "1" } }'
+          - '{ "@timestamp": "2022-01-01", "baz": "lazy", "a": "dog" }'
+  - is_true: errors
+  - match: { items.1.create._index: '/\.ds-logs-foobar-(\d{4}\.\d{2}\.\d{2}-)?000001/' }
+  - match: { items.1.create.status: 409 }
+  - match: { items.1.create.error.type: version_conflict_engine_exception}

server/src/main/java/org/elasticsearch/action/bulk/BulkOperation.java

@@ -44,6 +44,7 @@
 import org.elasticsearch.core.TimeValue;
 import org.elasticsearch.index.Index;
 import org.elasticsearch.index.IndexNotFoundException;
+import org.elasticsearch.index.engine.VersionConflictEngineException;
 import org.elasticsearch.index.shard.ShardId;
 import org.elasticsearch.indices.IndexClosedException;
 import org.elasticsearch.node.NodeClosedException;
@@ -478,15 +479,18 @@ private void completeShardOperation() {
             }
 
             private void processFailure(BulkItemRequest bulkItemRequest, Exception cause) {
-                var errorType = ElasticsearchException.getExceptionName(ExceptionsHelper.unwrapCause(cause));
+                var error = ExceptionsHelper.unwrapCause(cause);
+                var errorType = ElasticsearchException.getExceptionName(error);
                 DocWriteRequest<?> docWriteRequest = bulkItemRequest.request();
                 DataStream failureStoreCandidate = getRedirectTargetCandidate(docWriteRequest, getClusterState().metadata());
                 // If the candidate is not null, the BulkItemRequest targets a data stream, but we'll still have to check if
                 // it has the failure store enabled.
                 if (failureStoreCandidate != null) {
                     // Do not redirect documents to a failure store that were already headed to one.
                     var isFailureStoreDoc = docWriteRequest instanceof IndexRequest indexRequest && indexRequest.isWriteToFailureStore();
-                    if (isFailureStoreDoc == false && failureStoreCandidate.isFailureStoreEnabled()) {
+                    if (isFailureStoreDoc == false
+                        && failureStoreCandidate.isFailureStoreEnabled()
+                        && error instanceof VersionConflictEngineException == false) {
                         // Redirect to failure store.
                         maybeMarkFailureStoreForRollover(failureStoreCandidate);
                         addDocumentToRedirectRequests(bulkItemRequest, cause, failureStoreCandidate.getName());