use of org.apache.nifi.schema.access.SchemaNotFoundException in project nifi by apache.
the class RestSchemaRegistryClient method getSchema.
@Override
public RecordSchema getSchema(final int schemaId) throws IOException, SchemaNotFoundException {
// The Confluent Schema Registry's REST API does not provide us with the 'subject' (name) of a Schema given the ID.
// It will provide us only the text of the Schema itself. Therefore, in order to determine the name (which is required for
// a SchemaIdentifier), we must obtain a list of all Schema names, and then request each and every one of the schemas to determine
// if the ID requested matches the Schema's ID.
// To make this more efficient, we will cache a mapping of Schema Name to identifier, so that we can look this up more efficiently.
// Check if we have cached the Identifier to Name mapping
final String schemaPath = getSchemaPath(schemaId);
final JsonNode responseJson = fetchJsonResponse(schemaPath, "id " + schemaId);
final JsonNode subjectsJson = fetchJsonResponse("/subjects", "subjects array");
final ArrayNode subjectsList = (ArrayNode) subjectsJson;
JsonNode completeSchema = null;
for (JsonNode subject : subjectsList) {
try {
final String subjectName = subject.asText();
completeSchema = postJsonResponse("/subjects/" + subjectName, responseJson, "schema id: " + schemaId);
break;
} catch (SchemaNotFoundException e) {
continue;
}
}
if (completeSchema == null) {
throw new SchemaNotFoundException("could not get schema with id: " + schemaId);
}
final RecordSchema recordSchema = createRecordSchema(completeSchema);
return recordSchema;
}
use of org.apache.nifi.schema.access.SchemaNotFoundException in project nifi by apache.
the class RestSchemaRegistryClient method fetchJsonResponse.
private JsonNode fetchJsonResponse(final String pathSuffix, final String schemaDescription) throws SchemaNotFoundException, IOException {
String errorMessage = null;
for (final String baseUrl : baseUrls) {
final String path = getPath(pathSuffix);
final String trimmedBase = getTrimmedBase(baseUrl);
final String url = trimmedBase + path;
final WebTarget webTarget = client.target(url);
final Response response = webTarget.request().accept(MediaType.APPLICATION_JSON).get();
final int responseCode = response.getStatus();
if (responseCode == Response.Status.OK.getStatusCode()) {
final JsonNode responseJson = response.readEntity(JsonNode.class);
return responseJson;
}
if (responseCode == Response.Status.NOT_FOUND.getStatusCode()) {
throw new SchemaNotFoundException("Could not find Schema with " + schemaDescription + " from the Confluent Schema Registry located at " + baseUrl);
}
if (errorMessage == null) {
errorMessage = response.readEntity(String.class);
}
}
throw new IOException("Failed to retrieve Schema with " + schemaDescription + " from any of the Confluent Schema Registry URL's provided; failure response message: " + errorMessage);
}
use of org.apache.nifi.schema.access.SchemaNotFoundException in project nifi by apache.
the class RestSchemaRegistryClient method postJsonResponse.
private JsonNode postJsonResponse(final String pathSuffix, final JsonNode schema, final String schemaDescription) throws SchemaNotFoundException {
String errorMessage = null;
for (final String baseUrl : baseUrls) {
final String path = getPath(pathSuffix);
final String trimmedBase = getTrimmedBase(baseUrl);
final String url = trimmedBase + path;
final WebTarget builder = client.target(url);
final Response response = builder.request().accept(MediaType.APPLICATION_JSON).header(CONTENT_TYPE_HEADER, SCHEMA_REGISTRY_CONTENT_TYPE).post(Entity.json(schema.toString()));
final int responseCode = response.getStatus();
if (responseCode == Response.Status.NOT_FOUND.getStatusCode()) {
continue;
}
if (responseCode == Response.Status.OK.getStatusCode()) {
final JsonNode responseJson = response.readEntity(JsonNode.class);
return responseJson;
}
}
throw new SchemaNotFoundException("Failed to retrieve Schema with " + schemaDescription + " from any of the Confluent Schema Registry URL's provided; failure response message: " + errorMessage);
}
use of org.apache.nifi.schema.access.SchemaNotFoundException in project nifi by apache.
the class PutElasticsearchHttpRecord method onTrigger.
@Override
public void onTrigger(final ProcessContext context, final ProcessSession session) throws ProcessException {
FlowFile flowFile = session.get();
if (flowFile == null) {
return;
}
final RecordReaderFactory readerFactory = context.getProperty(RECORD_READER).asControllerService(RecordReaderFactory.class);
// Authentication
final String username = context.getProperty(USERNAME).evaluateAttributeExpressions(flowFile).getValue();
final String password = context.getProperty(PASSWORD).evaluateAttributeExpressions(flowFile).getValue();
OkHttpClient okHttpClient = getClient();
final ComponentLog logger = getLogger();
final String baseUrl = trimToEmpty(context.getProperty(ES_URL).evaluateAttributeExpressions().getValue());
HttpUrl.Builder urlBuilder = HttpUrl.parse(baseUrl).newBuilder().addPathSegment("_bulk");
// Find the user-added properties and set them as query parameters on the URL
for (Map.Entry<PropertyDescriptor, String> property : context.getProperties().entrySet()) {
PropertyDescriptor pd = property.getKey();
if (pd.isDynamic()) {
if (property.getValue() != null) {
urlBuilder = urlBuilder.addQueryParameter(pd.getName(), context.getProperty(pd).evaluateAttributeExpressions().getValue());
}
}
}
final URL url = urlBuilder.build().url();
final String index = context.getProperty(INDEX).evaluateAttributeExpressions(flowFile).getValue();
if (StringUtils.isEmpty(index)) {
logger.error("No value for index in for {}, transferring to failure", new Object[] { flowFile });
session.transfer(flowFile, REL_FAILURE);
return;
}
final String docType = context.getProperty(TYPE).evaluateAttributeExpressions(flowFile).getValue();
String indexOp = context.getProperty(INDEX_OP).evaluateAttributeExpressions(flowFile).getValue();
if (StringUtils.isEmpty(indexOp)) {
logger.error("No Index operation specified for {}, transferring to failure.", new Object[] { flowFile });
session.transfer(flowFile, REL_FAILURE);
return;
}
switch(indexOp.toLowerCase()) {
case "index":
case "update":
case "upsert":
case "delete":
break;
default:
logger.error("Index operation {} not supported for {}, transferring to failure.", new Object[] { indexOp, flowFile });
session.transfer(flowFile, REL_FAILURE);
return;
}
this.nullSuppression = context.getProperty(SUPPRESS_NULLS).getValue();
final String id_path = context.getProperty(ID_RECORD_PATH).evaluateAttributeExpressions(flowFile).getValue();
final RecordPath recordPath = StringUtils.isEmpty(id_path) ? null : recordPathCache.getCompiled(id_path);
final StringBuilder sb = new StringBuilder();
try (final InputStream in = session.read(flowFile);
final RecordReader reader = readerFactory.createRecordReader(flowFile, in, getLogger())) {
Record record;
while ((record = reader.nextRecord()) != null) {
final String id;
if (recordPath != null) {
Optional<FieldValue> idPathValue = recordPath.evaluate(record).getSelectedFields().findFirst();
if (!idPathValue.isPresent() || idPathValue.get().getValue() == null) {
throw new IdentifierNotFoundException("Identifier Record Path specified but no value was found, transferring {} to failure.");
}
id = idPathValue.get().getValue().toString();
} else {
id = null;
}
// a missing ID indicates one is to be auto-generated by Elasticsearch
if (id == null && !indexOp.equalsIgnoreCase("index")) {
throw new IdentifierNotFoundException("Index operation {} requires a valid identifier value from a flow file attribute, transferring to failure.");
}
final StringBuilder json = new StringBuilder();
ByteArrayOutputStream out = new ByteArrayOutputStream();
JsonGenerator generator = factory.createJsonGenerator(out);
writeRecord(record, record.getSchema(), generator);
generator.flush();
generator.close();
json.append(out.toString());
if (indexOp.equalsIgnoreCase("index")) {
sb.append("{\"index\": { \"_index\": \"");
sb.append(index);
sb.append("\", \"_type\": \"");
sb.append(docType);
sb.append("\"");
if (!StringUtils.isEmpty(id)) {
sb.append(", \"_id\": \"");
sb.append(id);
sb.append("\"");
}
sb.append("}}\n");
sb.append(json);
sb.append("\n");
} else if (indexOp.equalsIgnoreCase("upsert") || indexOp.equalsIgnoreCase("update")) {
sb.append("{\"update\": { \"_index\": \"");
sb.append(index);
sb.append("\", \"_type\": \"");
sb.append(docType);
sb.append("\", \"_id\": \"");
sb.append(id);
sb.append("\" }\n");
sb.append("{\"doc\": ");
sb.append(json);
sb.append(", \"doc_as_upsert\": ");
sb.append(indexOp.equalsIgnoreCase("upsert"));
sb.append(" }\n");
} else if (indexOp.equalsIgnoreCase("delete")) {
sb.append("{\"delete\": { \"_index\": \"");
sb.append(index);
sb.append("\", \"_type\": \"");
sb.append(docType);
sb.append("\", \"_id\": \"");
sb.append(id);
sb.append("\" }\n");
}
}
} catch (IdentifierNotFoundException infe) {
logger.error(infe.getMessage(), new Object[] { flowFile });
flowFile = session.penalize(flowFile);
session.transfer(flowFile, REL_FAILURE);
return;
} catch (final IOException | SchemaNotFoundException | MalformedRecordException e) {
logger.error("Could not parse incoming data", e);
flowFile = session.penalize(flowFile);
session.transfer(flowFile, REL_FAILURE);
return;
}
RequestBody requestBody = RequestBody.create(MediaType.parse("application/json"), sb.toString());
final Response getResponse;
try {
getResponse = sendRequestToElasticsearch(okHttpClient, url, username, password, "PUT", requestBody);
} catch (final Exception e) {
logger.error("Routing to {} due to exception: {}", new Object[] { REL_FAILURE.getName(), e }, e);
flowFile = session.penalize(flowFile);
session.transfer(flowFile, REL_FAILURE);
return;
}
final int statusCode = getResponse.code();
if (isSuccess(statusCode)) {
ResponseBody responseBody = getResponse.body();
try {
final byte[] bodyBytes = responseBody.bytes();
JsonNode responseJson = parseJsonResponse(new ByteArrayInputStream(bodyBytes));
boolean errors = responseJson.get("errors").asBoolean(false);
// ES has no rollback, so if errors occur, log them and route the whole flow file to failure
if (errors) {
ArrayNode itemNodeArray = (ArrayNode) responseJson.get("items");
if (itemNodeArray.size() > 0) {
// All items are returned whether they succeeded or failed, so iterate through the item array
// at the same time as the flow file list, moving each to success or failure accordingly,
// but only keep the first error for logging
String errorReason = null;
for (int i = itemNodeArray.size() - 1; i >= 0; i--) {
JsonNode itemNode = itemNodeArray.get(i);
int status = itemNode.findPath("status").asInt();
if (!isSuccess(status)) {
if (errorReason == null) {
// Use "result" if it is present; this happens for status codes like 404 Not Found, which may not have an error/reason
String reason = itemNode.findPath("//result").asText();
if (StringUtils.isEmpty(reason)) {
// If there was no result, we expect an error with a string description in the "reason" field
reason = itemNode.findPath("//error/reason").asText();
}
errorReason = reason;
logger.error("Failed to process {} due to {}, transferring to failure", new Object[] { flowFile, errorReason });
}
}
}
}
session.transfer(flowFile, REL_FAILURE);
} else {
session.transfer(flowFile, REL_SUCCESS);
session.getProvenanceReporter().send(flowFile, url.toString());
}
} catch (IOException ioe) {
// Something went wrong when parsing the response, log the error and route to failure
logger.error("Error parsing Bulk API response: {}", new Object[] { ioe.getMessage() }, ioe);
session.transfer(flowFile, REL_FAILURE);
context.yield();
}
} else if (statusCode / 100 == 5) {
// 5xx -> RETRY, but a server error might last a while, so yield
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 {
// 1xx, 3xx, 4xx, etc. -> NO RETRY
logger.warn("Elasticsearch returned code {} with message {}, transferring flow file to failure", new Object[] { statusCode, getResponse.message() });
session.transfer(flowFile, REL_FAILURE);
}
getResponse.close();
}
use of org.apache.nifi.schema.access.SchemaNotFoundException in project nifi by apache.
the class TestAvroSchemaRegistry method validateSchemaRegistrationFromrDynamicProperties.
@Test
public void validateSchemaRegistrationFromrDynamicProperties() throws Exception {
String schemaName = "fooSchema";
PropertyDescriptor fooSchema = new PropertyDescriptor.Builder().name(schemaName).dynamic(true).build();
String fooSchemaText = "{\"namespace\": \"example.avro\", " + "\"type\": \"record\", " + "\"name\": \"User\", " + "\"fields\": [ " + "{\"name\": \"name\", \"type\": [\"string\", \"null\"]}, " + "{\"name\": \"favorite_number\", \"type\": [\"int\", \"null\"]}, " + "{\"name\": \"foo\", \"type\": [\"int\", \"null\"]}, " + "{\"name\": \"favorite_color\", \"type\": [\"string\", \"null\"]} " + "]" + "}";
PropertyDescriptor barSchema = new PropertyDescriptor.Builder().name("barSchema").dynamic(false).build();
AvroSchemaRegistry delegate = new AvroSchemaRegistry();
delegate.onPropertyModified(fooSchema, null, fooSchemaText);
delegate.onPropertyModified(barSchema, null, "");
SchemaIdentifier schemaIdentifier = SchemaIdentifier.builder().name(schemaName).build();
RecordSchema locatedSchema = delegate.retrieveSchema(schemaIdentifier);
assertEquals(fooSchemaText, locatedSchema.getSchemaText().get());
try {
delegate.retrieveSchema(SchemaIdentifier.builder().name("barSchema").build());
Assert.fail("Expected a SchemaNotFoundException to be thrown but it was not");
} catch (final SchemaNotFoundException expected) {
}
}
Aggregations