Apicurio Registry 2.0.0!
by Eric Wittmann
Hey everyone. Happy Monday.
At the end of last week we released version 2.0.0.RC1 of Apicurio Registry
(product name: Red Hat Integration - Service Registry). This new major
version has a lot of changes and improvements. Here is quick list of cool
new stuff:
* Authentication via integration with Keycloak/RHSSO
* Support for groups of artifacts
* Updated Serdes classes for easier and more consistent configuration
* Support for the new CNCF Schema Registry API
* Multi-tenancy support! (This was mostly done to support MAS efforts)
* Option to store persistent registry data in a SQL database (postgresql
only)
* Hybrid storage implementation using Kafka (simple topic with log
compaction enabled) and an in-memory H2 database
* Improved artifact searching/filtering
* New REST client using only the Java 11 HTTP client
* Event sourcing (changes to artifacts in the registry can result in events
firing)
* And much more!
For more information about the changes, you can go here:
https://github.com/Apicurio/apicurio-registry/releases/tag/2.0.0.RC1
We will be doing a lot of testing and documenting over the next couple of
weeks before releasing 2.0.0.Final. If anyone is interested in giving the
new version a try, now is a great time! (you can report any bugs you find)
You can try it out immediately using our in-memory ( non-persistent )
version by doing this:
docker run -it -p 8080:8080 apicurio/apicurio-registry-mem:2.0.0.RC1
Then go to http://localhost:8080/ui or http://localhost:8080/apis to either
use the user interface or see a list of the APIs we support (respectively).
Thank you to everyone who contributed to this release! I think we should
be in a great position to support both our Red Hat Integration product as
well as our Managed Services efforts with this latest major version release.
--
Eric Wittmann
Principal Software Engineer - Apicurio - Red Hat
He / Him / His
eric.wittmann(a)redhat.com
2 months, 1 week
Junit test for Apicurio registry
by Janez Bindas
Hi all,
I would like to test Kafka workflow with Apicurio registry. I wasn't able
to registry the schema in memory. Can you tell me how to do it?
This is my test class:
@EmbeddedKafka
@ExtendWith(SpringExtension.class)
@TestInstance(TestInstance.Lifecycle.PER_CLASS)
public class KafkaTest {
private static final String TOPIC = "foobar";
@Autowired
private EmbeddedKafkaBroker embeddedKafkaBroker;
BlockingQueue<ConsumerRecord<FooBarKey, FooBarValue>> records;
KafkaMessageListenerContainer<FooBarKey, FooBarValue> container;
@BeforeEach
void setUp() {
Map<String, Object> configs = new
HashMap<>(KafkaTestUtils.consumerProps("consumer", "false",
embeddedKafkaBroker));
DefaultKafkaConsumerFactory<FooBarKey, FooBarValue>
consumerFactory = new DefaultKafkaConsumerFactory<>(configs, new
AvroKafkaDeserializer<>(), new AvroKafkaDeserializer<>());
ContainerProperties containerProperties = new
ContainerProperties(TOPIC);
container = new
KafkaMessageListenerContainer<>(consumerFactory, containerProperties);
records = new LinkedBlockingQueue<>();
container.setupMessageListener((MessageListener<FooBarKey,
FooBarValue>) records::add);
container.start();
ContainerTestUtils.waitForAssignment(container,
embeddedKafkaBroker.getPartitionsPerTopic());
}
@AfterEach
void tearDown() {
container.stop();
}
@Test
public void testKafka() throws InterruptedException {
Map<String, Object> configs = new
HashMap<>(KafkaTestUtils.producerProps(embeddedKafkaBroker));
Producer<Object, Object> producer = new
DefaultKafkaProducerFactory<>(configs, new AvroKafkaSerializer<>(),
new AvroKafkaSerializer<>()).createProducer();
FooBarKey fooBarKey = new FooBarKey();
fooBarKey.setId("33217330-1968-40c6-974a-d5aae70f3692");
FooBarValue fooBarValue = new FooBarValue();
fooBarValue.setBar("test");
fooBarValue.setFoo("Test2");
producer.send(new ProducerRecord<>(TOPIC, fooBarKey, fooBarValue));
producer.flush();
ConsumerRecord<FooBarKey, FooBarValue> singleRecord =
records.poll(100, TimeUnit.MILLISECONDS);
Assertions.assertThat(singleRecord).isNotNull();
}
}
avro file for key:
{
"type" : "record",
"name" : "FooBarKey",
"namespace" : "com.umwerk.vbank",
"doc" : "FooBar KEY schema",
"fields" : [ {
"name" : "id",
"type" : "string",
"doc" : "FooBar Id",
"logicalType" : "uuid"
} ]
}
avro file for value:
{
"type" : "record",
"name" : "FooBarValue",
"namespace" : "com.umwerk.vbank",
"doc" : "FooBar VALUE schema",
"fields" : [ {
"name" : "foo",
"type" : "string",
"doc" : "Test foo property"
}, {
"name" : "bar",
"type" : "string",
"doc" : "Test bar property"
} ]
}
Regards, Janez Bindas
2 years, 2 months
Java/REST API: Documentation of fields - globalId/contentId
by Jan Galinski
Hello,
we are starting evaluation of the apicurio registry. We will be using avro-schema and want to serialize to avro single-object-encoded (byteArray that contains a long-fingerprint to identify the schema).
This means that we will need a reliable way to query by the fingerprintId.
Can I use the globalId for this? When I upload the schema file via web-ui, the globalId is just based on a sequence, but it seems I would be able to update the metaData of this schema, so I could replace the globalId with an avro fingerprint.
Is globalId the correct field? What is the difference to the contentId?
Why are properties stored as key/value map, but queried as List<String>?
Thanks
Jan
3 years, 9 months
Mapping avro schema namespace/name to groupId/name
by Jan Galinski
Is it safe to assume that when my avro schema uses namespace="foo" and name="Bar" that it is a good idea to use groupId="foo" and name="Bar" as well when I upload the schema file?
If so ... couldn't the registry automatically derive namespace/name from an avro schema, so we do not risk typos when registering?
3 years, 9 months
