@Override
public void onTrigger(final ProcessContext context, final ProcessSession session)
throws ProcessException {
FlowFile flowFile = null;
if (context.hasIncomingConnection()) {
flowFile = session.get();
// If we have no FlowFile, and all incoming connections are self-loops then we can continue
// on.
// However, if we have no FlowFile and we have connections coming from other Processors, then
// we know that we should run only if we have a FlowFile.
if (flowFile == null && context.hasNonLoopConnection()) {
return;
}
}
OkHttpClient okHttpClient = getClient();
if (flowFile == null) {
flowFile = session.create();
}
final String index =
context.getProperty(INDEX).evaluateAttributeExpressions(flowFile).getValue();
final String docId =
context.getProperty(DOC_ID).evaluateAttributeExpressions(flowFile).getValue();
final String docType =
context.getProperty(TYPE).evaluateAttributeExpressions(flowFile).getValue();
final String fields =
context.getProperty(FIELDS).isSet()
? context.getProperty(FIELDS).evaluateAttributeExpressions(flowFile).getValue()
: null;
// Authentication
final String username =
context.getProperty(USERNAME).evaluateAttributeExpressions(flowFile).getValue();
final String password = context.getProperty(PASSWORD).evaluateAttributeExpressions().getValue();
final ComponentLog logger = getLogger();
Response getResponse = null;
try {
logger.debug("Fetching {}/{}/{} from Elasticsearch", new Object[] {index, docType, docId});
// read the url property from the context
final String urlstr =
StringUtils.trimToEmpty(
context.getProperty(ES_URL).evaluateAttributeExpressions().getValue());
final URL url = buildRequestURL(urlstr, docId, index, docType, fields);
final long startNanos = System.nanoTime();
getResponse = sendRequestToElasticsearch(okHttpClient, url, username, password, "GET", null);
final int statusCode = getResponse.code();
if (isSuccess(statusCode)) {
ResponseBody body = getResponse.body();
final byte[] bodyBytes = body.bytes();
JsonNode responseJson = parseJsonResponse(new ByteArrayInputStream(bodyBytes));
boolean found = responseJson.get("found").asBoolean(false);
String retrievedIndex = responseJson.get("_index").asText();
String retrievedType = responseJson.get("_type").asText();
String retrievedId = responseJson.get("_id").asText();
if (found) {
JsonNode source = responseJson.get("_source");
flowFile = session.putAttribute(flowFile, "filename", retrievedId);
flowFile = session.putAttribute(flowFile, "es.index", retrievedIndex);
flowFile = session.putAttribute(flowFile, "es.type", retrievedType);
if (source != null) {
flowFile =
session.write(
flowFile,
out -> {
out.write(source.toString().getBytes());
});
}
logger.debug("Elasticsearch document " + retrievedId + " fetched, routing to success");
// emit provenance event
final long millis = TimeUnit.NANOSECONDS.toMillis(System.nanoTime() - startNanos);
if (context.hasNonLoopConnection()) {
session.getProvenanceReporter().fetch(flowFile, url.toExternalForm(), millis);
} else {
session.getProvenanceReporter().receive(flowFile, url.toExternalForm(), millis);
}
session.transfer(flowFile, REL_SUCCESS);
} else {
logger.warn(
"Failed to read {}/{}/{} from Elasticsearch: Document not found",
new Object[] {index, docType, docId});
// We couldn't find the document, so send it to "not found"
session.transfer(flowFile, REL_NOT_FOUND);
}
} else {
if (statusCode == 404) {
logger.warn(
"Failed to read {}/{}/{} from Elasticsearch: Document not found",
new Object[] {index, docType, docId});
// We couldn't find the document, so penalize it and send it to "not found"
session.transfer(flowFile, REL_NOT_FOUND);
} else {
// 5xx -> RETRY, but a server error might last a while, so yield
if (statusCode / 100 == 5) {
logger.warn(
"Elasticsearch returned code {} with message {}, transferring flow file to retry. This is likely a server problem, yielding...",
new Object[] {statusCode, getResponse.message()});
session.transfer(flowFile, REL_RETRY);
context.yield();
} else if (context.hasIncomingConnection()) { // 1xx, 3xx, 4xx -> NO RETRY
logger.warn(
"Elasticsearch returned code {} with message {}, transferring flow file to failure",
new Object[] {statusCode, getResponse.message()});
session.transfer(flowFile, REL_FAILURE);
} else {
logger.warn(
"Elasticsearch returned code {} with message {}",
new Object[] {statusCode, getResponse.message()});
session.remove(flowFile);
}
}
}
} catch (IOException ioe) {
logger.error(
"Failed to read from Elasticsearch due to {}, this may indicate an error in configuration "
+ "(hosts, username/password, etc.). Routing to retry",
new Object[] {ioe.getLocalizedMessage()},
ioe);
if (context.hasIncomingConnection()) {
session.transfer(flowFile, REL_RETRY);
} else {
session.remove(flowFile);
}
context.yield();
} catch (Exception e) {
logger.error(
"Failed to read {} from Elasticsearch due to {}",
new Object[] {flowFile, e.getLocalizedMessage()},
e);
if (context.hasIncomingConnection()) {
session.transfer(flowFile, REL_FAILURE);
} else {
session.remove(flowFile);
}
context.yield();
} finally {
if (getResponse != null) {
getResponse.close();
}
}
}
