apicurio / apicurio-registry Goto Github PK

An API/Schema registry - stores APIs and Schemas.

Home Page: https://www.apicur.io/registry/

License: Apache License 2.0

Java 71.03% Shell 0.14% HTML 0.08% JavaScript 0.14% TypeScript 7.68% CSS 0.55% Makefile 0.02% Python 0.15% Dockerfile 0.04% Go 20.15% Handlebars 0.03%

apicurio-registry's Issues

Enable metrics for the app

Use whatever standard metrics feature Quarkus has - presumably MP metrics.

Validation Rule: AsyncAPI

Implement the Validation Rule for AsyncAPI content.

Create Indexing Client for ISPN storage

Implement the JAX-RS Global Rules interface

Has anyone managed to get their own setup working against corporate security with SAML? We're trying to have SAML for all applications and have an in house version of apicurio for our development teams to use.

I attempted to wire the code directly to SAML and bypass keycloak but was unable to figure out how JBOSS security is wired up in this area as its inside the server and no matter how many google searches I took no luck finding anyone else doing this either.

Seeing the code is setup for Keycloak now wondering if keycloak can use saml as an identify provider but maybe someone else has done this before.

Deploy: Use Quarkus configuration & conf. profiles

Use Quarkus's configuration system to create a set of profiles for various kind of deployment scenarios, including selection of the persistence layer implementation and logging.

Profiles could be e.g.: dev (h2 DB), single (PostgreSQL) cluster (Infinispan), kafka, ...

Resources:

https://quarkus.io/guides/application-configuration-guide#configuration-profiles

Compatibility Rule: Kafka Connect Schema

Document: content-types for supported formats

We're going to be storing artifacts of the following formats in the registry:

JSON Schema
Avro
Protobuff
OpenAPI
AsyncAPI

Most of these are in JSON format (or YAML which can easily be converted). But at least Protobuff is not. So there is an open question about how to handle the different content types when pushing content into the registry via the API and then returning it later.

The first step is to identify the actual content types for each format. And then additionally how to handle that in the API itself. In order to make it as simple as possible, we could try to figure out the type based on the following (in order):

custom Request header e.g. X-Registry-Type
check the Request's Content-Type for something that disambiguates the content. For example the request Content-Type might be application/json+openapi
try to figure out the type from the content itself. This will work well for OpenAPI and AsyncAPI and probably Protobuff, but I'm not sure what else.

Downloading / Publishing YAML removes quotes from strings

When downloading or publishing an project api specification, we're seeing previously quoted values being outputted without quotes.

This is in turn leading to a bug with the fields being incorrectly rendered in tooling such as redoc — which typically converts the YAML into a JSON object

A repeatable example can be seen over at https://onlineyamltools.com/convert-yaml-to-json using the following example payloads...

Quoted Date String

Priority Order request:
  value:
    order:
      courier_type: priority
      delivery_address: 1 Union Street
      delivery_postcode: "2009"
      delivery_state: NSW
      delivery_suburb: Pyrmont
      authority_to_leave: "Yes"
      delivery_date: "2016-07-26"
      delivery_window: 16:00-19:00
      parcel_attributes:
      - qty: 1
        weight: 2.1

— delivery_date results in displaying as "2016-07-26"

Unquoted Date String

Priority Order request:
  value:
    order:
      courier_type: priority
      delivery_address: 1 Union Street
      delivery_postcode: "2009"
      delivery_state: NSW
      delivery_suburb: Pyrmont
      authority_to_leave: "Yes"
      delivery_date: 2016-07-26
      delivery_window: 16:00-19:00
      parcel_attributes:
      - qty: 1
        weight: 2.1

— delivery_date results in displaying as "2016-07-26T00:00:00.000Z"

Determine artifact type based on content

If the user does not provide the artifact type via a request header, then the registry needs to probe the content to try and determine the type. This amounts to attempting to parse the content as JSON, then as protobuf. If it's JSON, then further analysis is needed to determine if it's an OpenAPI, AsyncAPI, JSON Schema, or Avro artifact.

Create an image using CEKit

https://github.com/cekit/cekit

Add disable/enable states to Artifacts (including additional REST API operations)

Deploy: Create OpenShift template

Storage Implementation: in-memory

An in-memory implementation of the storage interface.

Implement serdes support for Kafka Connect schemas

Consult with Gunnar Morling on requirements and implementation details.

Validation Rule: Protobuff

Implement the Validation Rule for Protobuff content. Rules will need to be defined for this.

Storage Implementation: rdbms

A relational database (sql) implementation of the storage interface. Can use hibernate or straight jdbc (IMO).

Some resources:

Hibernate + Quarkus: https://quarkus.io/guides/hibernate-orm-guide
Hibernate and Panache + Quarkus: https://quarkus.io/guides/hibernate-orm-panache-guide

Validation Rule: Avro

Implement the Validation Rule for Avro content. Rules will need to be defined for this.

Maintenance: Maven build checkup

Go over the POM files, profiles, tests etc. in case there is some maintenance work needed. At the very least, we need to make sure that mvn clean install will build, test, and package all of the supported storage variants. Right now the variants are all separated out into profiles, because the tests for "streams", "kafka", and "infinispan" will all fail because they all depend on external systems to function. We need to work on standing up those systems automatically so that tests can be run successfully.

Deploy: Create OpenShift operator

Create an ansible-based operator for installing the registry on OpenShift.

Add support for GraphQL artifacts

Validation Rule: JSON Schema

Implement the Validation Rule for JSON Schema content. Rules will need to be defined for this.

Implement serdes for JSON Schema

API: Design the registry API

Create a REST API for the registry that is more generic than the confluent compatible API - it should handle multiple content types and support rules instead of just compatibility.

Validity Rule: Kafka Connect Schema

Create a utility to migrate data from confluent schema registry to apicurio registry

Should help with migration from one system to the other.

Enable healthcheck for the app

Use the recommended Quarkus approach to enabling app healthchecking. I presume this is MP health.

Compatibility Rule: GraphQL

Upgrade to Quarkus 0.26.1

Quarkus 0.26.1 has already changed enough that this project fails to compile against it. :(
Following bleeding edge projects is hard.

Compatibility Rule: OpenAPI

Implement the compatibility rule for OpenAPI schemas. We'll need to really analyze the avro version to implement similar logic for OpenAPI.

Create Indexing Client for JPA storage

Storage Implementation: kafka

A kafka implementation of the storage interface.

Jersey client cannot parse replies from RESTEasy server

If Avro converter is used in Kafka COnnect then Jersey REST client must/should be used. The problem is that when an atifact like /artifacts/dbserver1-key is created the server anwers with

{"createdOn":1579862056239,"modifiedOn":1579864479545,"id":"dbserver1-key","version":3,"type":"AVRO","globalId":3}

Date fields are serialized as epoch milliseconds.

When the client tries to parse them it gets

connect_1    | SEVERE: Unable to deserialize property 'createdOn' because of: Error parsing class java.util.Date from value: 1579862056239. Check your @JsonbDateFormat has all time units for class java.util.Date type, or consider using org.eclipse.yasson.YassonProperties#ZERO_TIME_PARSE_DEFAULTING.
connect_1    | 2020-01-24 11:14:39,550 INFO   ||  WorkerSourceTask{id=inventory-connector-0} Committing offsets   [org.apache.kafka.connect.runtime.WorkerSourceTask]
connect_1    | 2020-01-24 11:14:39,550 INFO   ||  WorkerSourceTask{id=inventory-connector-0} flushing 0 outstanding messages for offset commit   [org.apache.kafka.connect.runtime.WorkerSourceTask]
connect_1    | 2020-01-24 11:14:39,550 ERROR  ||  WorkerSourceTask{id=inventory-connector-0} Task threw an uncaught and unrecoverable exception   [org.apache.kafka.connect.runtime.WorkerTask]
connect_1    | org.apache.kafka.connect.errors.ConnectException: Tolerance exceeded in error handler
connect_1    | 	at org.apache.kafka.connect.runtime.errors.RetryWithToleranceOperator.execAndHandleError(RetryWithToleranceOperator.java:178)
connect_1    | 	at org.apache.kafka.connect.runtime.errors.RetryWithToleranceOperator.execute(RetryWithToleranceOperator.java:104)
connect_1    | 	at org.apache.kafka.connect.runtime.WorkerSourceTask.convertTransformedRecord(WorkerSourceTask.java:287)
connect_1    | 	at org.apache.kafka.connect.runtime.WorkerSourceTask.sendRecords(WorkerSourceTask.java:316)
connect_1    | 	at org.apache.kafka.connect.runtime.WorkerSourceTask.execute(WorkerSourceTask.java:240)
connect_1    | 	at org.apache.kafka.connect.runtime.WorkerTask.doRun(WorkerTask.java:177)
connect_1    | 	at org.apache.kafka.connect.runtime.WorkerTask.run(WorkerTask.java:227)
connect_1    | 	at java.base/java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:515)
connect_1    | 	at java.base/java.util.concurrent.FutureTask.run(FutureTask.java:264)
connect_1    | 	at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1128)
connect_1    | 	at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:628)
connect_1    | 	at java.base/java.lang.Thread.run(Thread.java:834)
connect_1    | Caused by: javax.ws.rs.ProcessingException: Error deserializing object from entity stream.
connect_1    | 	at org.glassfish.jersey.jsonb.internal.JsonBindingProvider.readFrom(JsonBindingProvider.java:77)
connect_1    | 	at org.glassfish.jersey.message.internal.ReaderInterceptorExecutor$TerminalReaderInterceptor.invokeReadFrom(ReaderInterceptorExecutor.java:233)
connect_1    | 	at org.glassfish.jersey.message.internal.ReaderInterceptorExecutor$TerminalReaderInterceptor.aroundReadFrom(ReaderInterceptorExecutor.java:212)
connect_1    | 	at org.glassfish.jersey.message.internal.ReaderInterceptorExecutor.proceed(ReaderInterceptorExecutor.java:132)
connect_1    | 	at org.glassfish.jersey.message.internal.MessageBodyFactory.readFrom(MessageBodyFactory.java:1071)
connect_1    | 	at org.glassfish.jersey.message.internal.InboundMessageContext.readEntity(InboundMessageContext.java:850)
connect_1    | 	at org.glassfish.jersey.message.internal.InboundMessageContext.readEntity(InboundMessageContext.java:810)
connect_1    | 	at org.glassfish.jersey.client.ClientResponse.readEntity(ClientResponse.java:339)
connect_1    | 	at org.glassfish.jersey.client.InboundJaxrsResponse$2.call(InboundJaxrsResponse.java:102)
connect_1    | 	at org.glassfish.jersey.internal.Errors.process(Errors.java:292)
connect_1    | 	at org.glassfish.jersey.internal.Errors.process(Errors.java:274)
connect_1    | 	at org.glassfish.jersey.internal.Errors.process(Errors.java:205)
connect_1    | 	at org.glassfish.jersey.process.internal.RequestScope.runInScope(RequestScope.java:365)
connect_1    | 	at org.glassfish.jersey.client.InboundJaxrsResponse.runInScopeIfPossible(InboundJaxrsResponse.java:240)
connect_1    | 	at org.glassfish.jersey.client.InboundJaxrsResponse.readEntity(InboundJaxrsResponse.java:99)
connect_1    | 	at org.glassfish.jersey.microprofile.restclient.MethodModel.lambda$asynchronousCall$5(MethodModel.java:266)
connect_1    | 	at java.base/java.util.concurrent.CompletableFuture$UniAccept.tryFire(CompletableFuture.java:714)
connect_1    | 	at java.base/java.util.concurrent.CompletableFuture.postComplete(CompletableFuture.java:506)
connect_1    | 	at java.base/java.util.concurrent.CompletableFuture.complete(CompletableFuture.java:2073)
connect_1    | 	at org.glassfish.jersey.client.JerseyInvocation$1.completed(JerseyInvocation.java:789)
connect_1    | 	at org.glassfish.jersey.client.ClientRuntime.processResponse(ClientRuntime.java:203)
connect_1    | 	at org.glassfish.jersey.client.ClientRuntime.access$200(ClientRuntime.java:61)
connect_1    | 	at org.glassfish.jersey.client.ClientRuntime$2.lambda$response$0(ClientRuntime.java:154)
connect_1    | 	at org.glassfish.jersey.internal.Errors$1.call(Errors.java:248)
connect_1    | 	at org.glassfish.jersey.internal.Errors$1.call(Errors.java:244)
connect_1    | 	at org.glassfish.jersey.internal.Errors.process(Errors.java:292)
connect_1    | 	at org.glassfish.jersey.internal.Errors.process(Errors.java:274)
connect_1    | 	at org.glassfish.jersey.internal.Errors.process(Errors.java:244)
connect_1    | 	at org.glassfish.jersey.process.internal.RequestScope.runInScope(RequestScope.java:288)
connect_1    | 	at org.glassfish.jersey.client.ClientRuntime$2.response(ClientRuntime.java:154)
connect_1    | 	at org.glassfish.jersey.client.internal.HttpUrlConnector.apply(HttpUrlConnector.java:268)
connect_1    | 	at org.glassfish.jersey.client.ClientRuntime.lambda$null$3(ClientRuntime.java:163)
connect_1    | 	at org.glassfish.jersey.internal.Errors$1.call(Errors.java:248)
connect_1    | 	at org.glassfish.jersey.internal.Errors$1.call(Errors.java:244)
connect_1    | 	at org.glassfish.jersey.internal.Errors.process(Errors.java:292)
connect_1    | 	at org.glassfish.jersey.internal.Errors.process(Errors.java:274)
connect_1    | 	at org.glassfish.jersey.internal.Errors.process(Errors.java:244)
connect_1    | 	at org.glassfish.jersey.process.internal.RequestScope.runInScope(RequestScope.java:288)
connect_1    | 	at org.glassfish.jersey.client.ClientRuntime.lambda$createRunnableForAsyncProcessing$4(ClientRuntime.java:139)
connect_1    | 	at org.glassfish.jersey.microprofile.restclient.ExecutorServiceWrapper.lambda$wrap$1(ExecutorServiceWrapper.java:124)
connect_1    | 	... 5 more
connect_1    | Caused by: javax.json.bind.JsonbException: Unable to deserialize property 'createdOn' because of: Error parsing class java.util.Date from value: 1579862056239. Check your @JsonbDateFormat has all time units for class java.util.Date type, or consider using org.eclipse.yasson.YassonProperties#ZERO_TIME_PARSE_DEFAULTING.
connect_1    | 	at org.eclipse.yasson.internal.serializer.AbstractContainerDeserializer.deserializeInternal(AbstractContainerDeserializer.java:90)
connect_1    | 	at org.eclipse.yasson.internal.serializer.AbstractContainerDeserializer.deserialize(AbstractContainerDeserializer.java:60)
connect_1    | 	at org.eclipse.yasson.internal.Unmarshaller.deserializeItem(Unmarshaller.java:68)
connect_1    | 	at org.eclipse.yasson.internal.Unmarshaller.deserialize(Unmarshaller.java:54)
connect_1    | 	at org.eclipse.yasson.internal.JsonBinding.deserialize(JsonBinding.java:53)
connect_1    | 	at org.eclipse.yasson.internal.JsonBinding.fromJson(JsonBinding.java:93)
connect_1    | 	at org.glassfish.jersey.jsonb.internal.JsonBindingProvider.readFrom(JsonBindingProvider.java:75)
connect_1    | 	... 44 more
connect_1    | Caused by: javax.json.bind.JsonbException: Error parsing class java.util.Date from value: 1579862056239. Check your @JsonbDateFormat has all time units for class java.util.Date type, or consider using org.eclipse.yasson.YassonProperties#ZERO_TIME_PARSE_DEFAULTING.
connect_1    | 	at org.eclipse.yasson.internal.serializer.AbstractDateTimeDeserializer.deserialize(AbstractDateTimeDeserializer.java:71)
connect_1    | 	at org.eclipse.yasson.internal.serializer.AbstractValueTypeDeserializer.deserialize(AbstractValueTypeDeserializer.java:64)
connect_1    | 	at org.eclipse.yasson.internal.serializer.ObjectDeserializer.deserializeNext(ObjectDeserializer.java:174)
connect_1    | 	at org.eclipse.yasson.internal.serializer.AbstractContainerDeserializer.deserializeInternal(AbstractContainerDeserializer.java:84)
connect_1    | 	... 50 more
connect_1    | Caused by: java.time.format.DateTimeParseException: Text '1579862056239' could not be parsed at index 0
connect_1    | 	at java.base/java.time.format.DateTimeFormatter.parseResolved0(DateTimeFormatter.java:2046)
connect_1    | 	at java.base/java.time.format.DateTimeFormatter.parse(DateTimeFormatter.java:1948)
connect_1    | 	at java.base/java.time.ZonedDateTime.parse(ZonedDateTime.java:598)
connect_1    | 	at org.eclipse.yasson.internal.serializer.DateTypeDeserializer.parseWithOrWithoutZone(DateTypeDeserializer.java:83)
connect_1    | 	at org.eclipse.yasson.internal.serializer.DateTypeDeserializer.parseDefault(DateTypeDeserializer.java:54)
connect_1    | 	at org.eclipse.yasson.internal.serializer.DateTypeDeserializer.parseDefault(DateTypeDeserializer.java:34)
connect_1    | 	at org.eclipse.yasson.internal.serializer.AbstractDateTimeDeserializer.deserialize(AbstractDateTimeDeserializer.java:69)
connect_1    | 	... 53 more

The clients expects it to be a string formatted value.

Add support for "deprecate" action

Improve logging

We don't really have any useful logging in registry or in the serdes right now. We should go through and add appropriate logging (info/debug/tracing).

Compatibility Rule: protobuff

Implement the compatibility rule for protobuff schemas. We'll need to really analyze the avro version to implement similar logic for protobuff.

Implement the JAX-RS Artifacts interface

Create an implementation of the "Artifacts" interface.

Compatibility Rule: AsyncAPI

Implement the compatibility rule for AsyncAPI schemas. We'll need to really analyze the avro version to implement similar logic for AsyncAPI.

Add support for Kafka Connect Schema artifacts

Kafka Connect uses a schema format that is not JSON Schema but is more similar to Avro (but is not Avro). We should add artifact type support for this type.

Validation Rule interface

We need a simple, common implementation that all validation rules will need to implement. This might be as simple as passing in the content and returning either a pass/fail or perhaps a list of errors (with empty list indicating PASS).

Examples of rules that we will implement:

Compatibility Rule (config and logic ported from Confluent avro compatibility feature)
Validation Rule (checks that the artifact conforms to configured company standards)

Compatibility Rule: avro

Implement the compatibility rule for avro schemas. This can be lifted from Perspicuus most likely.

Improve error handling in serdes

If the user does not send the right thing to the topic when sending a message, the serdes Serializer doesn't throw a sensible error. It just throws a "404 not found" error. I think this is because it's trying to find the Schema in the registry even when the message is e.g. a simple type. It would better if the serializer could detect that the data being sent isn't appropriate for serialization (e.g. it's just a simple type) and throw an appropriate error.

in-memory
rdbms
kafka

/*
 * Copyright 2019 Red Hat
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *      http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

Duplicate update in JPA?

@jsenko why do we do duplicate persist here?
First with nextVersion, then with null as version ... ?!

apicurio-registry/storage/jpa/src/main/java/io/apicurio/registry/storage/impl/jpa/JPARegistryStorage.java

Line 215 in e74d5fa

.persistUpdate(entityManager, artifactId, null)

apicurio / apicurio-registry Goto Github PK

apicurio-registry's Issues

Recommend Projects

Recommend Topics

Recommend Org