Remove failure_store field name from Data Stream serialization

[jbaiera]

Currently we serialize if a data stream has the failure store explicitly enabled when writing a data stream to XContent and to a byte stream. This is redundant because we always send the data stream options for the data stream which already have this setting inside of them. This is a vestigial field from early in failure store's dev cycle that has yet to be cleared up. Now that the feature is going into GA - logic that uses values provided to this field has be cleared up to ignore it. It doesn't look like the field is currently written in XContent. We should look toward clearing it out.

elasticsearch/server/src/main/java/org/elasticsearch/cluster/metadata/DataStream.java

Lines 1253 to 1256 in d928d1a

	if (out.getTransportVersion()
	.between(DataStream.ADDED_FAILURE_STORE_TRANSPORT_VERSION, DataStream.ADD_DATA_STREAM_OPTIONS_VERSION)) {
	out.writeBoolean(isFailureStoreExplicitlyEnabled());
	}

elasticsearch/server/src/main/java/org/elasticsearch/cluster/metadata/DataStream.java

Line 1286 in d928d1a

public static final ParseField FAILURE_STORE_FIELD = new ParseField("failure_store");

elasticsearch/server/src/main/java/org/elasticsearch/cluster/metadata/DataStream.java

Lines 1366 to 1367 in d928d1a

	// Should be removed after backport
	PARSER.declareBoolean(ConstructingObjectParser.optionalConstructorArg(), FAILURE_STORE_FIELD);

[elasticsearchmachine]

Pinging @elastic/es-data-management (Team:Data Management)