Camel Rebirth with Subsecond Experiences

This post was originally published here. Check out my new Kubernetes Patterns book and  follow me @bibryam for future blog posts.

A look at the past decade

The integration space is in constant change. There are many open source projects and closed source technologies that have not passed the test of time and disappeared from the middleware stacks for good. After a decade, Apache Camel is still here and becoming even stronger for the next decade of integration.

Gang of Four for integration

Apache Camel started life as an implementation of the Enterprise Integration Patterns (EIP) book. Today, these patterns are the equivalent of the object-oriented Gang of Four Design Patterns but for messaging and integration domain. They are agnostic of programing language, platform, architecture, and provide a universal language, notation and description of the forces around fundamental messaging primitives.

But Camel community did not stop with these patterns, and kept evolving and adding newer patterns from SOA, Microservices, Cloud Native and Serverless paradigms. As a result, Camel turned into a generic pattern based integration framework suitable for multiple architectures.

Universal library consumption model

While the patterns gave the initial spark to Camel, its endpoints quickly became popular and turned into a universal protocol for using Java based integration libraries as connectors. Today, there are hundreds of Java libraries that can be used as Camel connectors using the Camel endpoint notation. It takes a while to realize that Camel can also be used without the EIPs and the routing engine. It can act as a connector framework where all libraries are consumed as universal URIs without a need for understanding the library specific factories and configurations that vary widely across Java libraries.

The right level of abstraction

When you talk to developers who have not used Camel in anger before, they will tell you that it is possible to do integration without Camel. And they are right about the 80% of the easy use cases, but not for the remaining 20% of the cases that can turn a project into a multi-year frustrating experience. What they do not realize yet is that without Camel, there are multiple manual ways of doing the same thing, but none are validated by the experience of hundreds of open source developers. And if you fast-forward a few years, 10s of different systems to integrate with, and 10s of developers coming and going, 100s of microservices, an integration project can quickly turn into a bespoke home-grown framework that nobody wants to work on. Doing integration is easy, but doing good integration that will evolve and grow for many years, by many teams, is the hardest part. Camel addresses this challenge with universal patterns and connectors, combined with integration focused DSLs, that have passed the test of time. Next time, if you think you don't need Camel, your are either thinking for short term gains, or you are not realizing yet how complex integration can become.

Embracing change

It takes only a couple of painful experiences in large integration projects to start appreciating Camel. But Camel is not great only because it was started by and built on the works of great minds, it is great because it evolves thanks to the world's knowledge, shared through the open source model and its networking effects. Camel started as the routing layer in ESBs during SOA period with a lot of focus on XML, WS, JBI, OSGI etc, but then it quickly adapted for REST, Swagger, Circuit breakers, SAGAs, and Spring Boot in the Microservices era. And the world has not stopped there, with Docker and Kubernetes, and now Serverless architecture, Camel keeps embracing the change. That is because Camel is written for integrating changing environments, and Camel itself grows and shines on change. Camel is a change enabling library for integration.

Behind the scene engine

One of the Camel secret sauces is that it is a non-intrusive, unopinionated, small (5MB and getting smaller) integration library without any affinity where and how you consume it. If you notice, this is the opposite of an ESB as commonly Camel is confused with because of its extensive capabilities. Over the years, Camel has been used as the internal engine powering projects such as:

• Apache ServiceMix ESB
• Apache ActiveMQ
• Talend ESB
• JBoss Switchyard
• JBoss Fuse Service Works
• Red Hat Fuse
• Fuse Online/Syndesis 
• And many other frameworks mentioned here.

You can use Camel standalone, embed it with Apache Tomcat, with Spring Boot starters, JBoss WildFly, Apache Karaf, Vert.x, Quarkus, you name it. Camel doesn't care and it will bring superpowers to your project every time.

Looking to the future

Nobody can tell how the ideal integration stack will look like in a decade, nor can I. But I will tell you about two novelties coming into Apache Camel now (and to Red Hat Fuse later), and why they will have a noticeable positive effect for the developers and the business. I call these changes as subsecond deployment and subsecond startup of Camel applications.

Subsecond deployments to Kubernetes

There was a time when cloud-native meant different technologies. Today, after a few years of natural selection and consolidation in the industry, cloud-native means applications created specifically for Kubernetes and its ecosystem of projects around CNCF. Even with this definition, there are many shades of cloud-native, from running a monolithic non-scalable application in a container, to triggering a function that is fully embracing the cloud-native development and management practices. The Camel community has realized that Kubernetes is the next generation application runtime, and it is steadily working on making Camel a Kubernetes native integration engine. The same way Camel is a first-class citizen in OSGI containers, JEE application servers, other fat-jar runtimes, Camel is becoming a first-class citizen on Kubernetes, integrating deeply and benefiting from the resiliency and scalability offered by the platform.

Here are a few of the many enhancement efforts going on in this direction:

• Deeper Kubernetes integration - Kubernetes API connector, full health-check API implementation for Camel subsystems, service discovery through a new ServiceCall EIP, configuration management using ConfigMaps. Then a set of application patterns with special handling on Kubernetes such as: clustered singleton routes, scalable XA transactions (because sometimes, you have to), SAGA pattern implementation, etc.
• Cloud-native integrations - support for other cloud-native projects such as exposing Camel metrics for Prometheus, tracing Camel routes through Jaeger, JSON formatted logging for log aggregation, etc.
• Immutable runtimes - whether you use the minimalist immutable Karaf packaging, or Spring Boot, Camel is a first class citizen ready to put in a container image. There are also Spring Boot starter implementations for all Camel connectors, integration with routes, properties, converters, and whatnot.
• Camel 3 - is a fact and actively progressing. A big theme for Camel 3 is to make it more modular, smaller, with faster startup time, reactive, non-blocking and triple awesome. This is the groundwork needed to restructure Camel for the future cloud workloads.
• Knative integration - Knative is an effort started by Google in order to bring some order and standardization in the serverless world dominated by Amazon Lambda. And Camel is among the projects that integrate with Knative primitives from early days and enhances the Knative ecosystem with hundreds of connectors acting as generic event sources.  
• And here is a real game-changer initiative: Camel-K (a.k.a deep Kubernetes integration for Camel) - we have seen that Camel is typically embedded into the latest modern runtime where it acts as the developer-friendly integration engine behind the scene. The same way Camel used to benefit from JEE services in the past for hot-deployment, configuration management, transaction management, etc, today Camel-K allows Camel runtime to benefit from Kubernetes features for high-availability, resiliency, self-healing, auto-scaling, and basically distributed application management in general. The way Camel-K achieves this is through a CLI and an Operator where the latter is able to understand the Camel applications, its build time dependencies, runtime needs, and make intelligent choices from the underlying Kubernetes platform and its additional capabilities (from Knative, Istio, Openshift and others in the future). It can automate everything on the cluster such as picking the best suited container image, runtime management model and update them when needed. The CLI can automate the tasks that are on the developer machine such as observing the code changes and streaming those to the Kubernetes cluster, and printing the logs from the running Pods. 

Camel route auto-deployment to Kubernetes with Camel-K

Camel-K operator understands two domains: Kubernetes and Camel. By combining knowledge of both areas, it can automate tasks that usually require a human operator.
The really powerful part is that, with Camel-K, a Camel route can be built and deployed from source code to a running Camel route on Kubernetes in less than a second!

Time to deploy and run a Camel integration(in seconds)

 

Forget about making a coffee, or even having a sip while building and deploying a Camel route with Camel K. As soon as you make changes to your source code and open a browser, the Camel route will be running in Kubernetes. This has a noticeable impact on the way the developers write Camel code, compile, drink coffee, deploy and test. Apart from changing the development practices and habits, this toolset will significantly reduce the development cycles which would be noticed by the business stakeholders too. For live demonstration, check out the following awesome video from Fuse engineers working on Camel-K project.

Subsecond startups of Camel applications

A typical enterprise integration landscape is composed of stateless services, stateful services, clustered applications, batch jobs, file transfers, messaging, real time integrations, and may be even blockchain based business processes. To that mix, today, we also have to add serverless workloads as well, which is best suited for event driven workloads. Historically, the heavy and slow Java runtime had significant drawbacks compared Go, Javascript and other light runtimes in the serverless space. That is one of the main motivations for Oracle to create GraalVM/Substrate VM. Substrate VM is a framework that enables ahead-of-time (AOT) compilation of Java applications into native executables that are light and fast. Then a recent effort by Red Hat led to the creation of Quarkus project which improves further the resource consumptions, startup and response times of Java applications mind-blowingly (a term not-used lightly here).

Supersonic Subatomic Java with Quarkus

 

As you can see from the metrics above, Quarkus combined with SubstrateVM is not a gradual evolution. It is a mutation, and a revolutionary jump that suddenly changes the perspectives on the Java’s footprint and speed in the cloud native workloads. It makes Java friendly for serverless architecture. Considering the huge Java ecosystem composed of developers and libraries, it even turns Java into the best suited language for Serverless applications. And Camel combined with Quarkus, the best placed integration library in this space.

Summary

With the explosion of Microservices architecture, the number of services has increased tenfold which gave birth to Kubernetes-enabled architectures. These architectures are highly dynamic in nature, and most powerful with light and fast runtimes that enable instant scale up and higher deployment density.

Camel is the framework to fill the space between disparate systems and services. It offers data consistency guarantees, reliable communication, failover, failure detection and recovery, and so on, in a way that makes developers productive. Now, imagine the same powerful Apache Camel based integration in 2020 that: deploys to Kubernetes in 20ms; starts up in 20ms; requires 20MB memory, consumes 20MB on the disk... That is regardless whether it runs as a stateless application in a container, or as a function on KNative. That is 100x faster deployments to Kubernetes, 100x faster startup time, 10x less resource consumption allowing real-time scale-up, scale-down, and scale to zero. That is a change that the developers will notice during development, users will notice when using the system, and the business will notice on the infrastructure cost and overall delivery velocity. That is the real cloud-native era we have been waiting for.

The next integration evolution - blockchain

Below is the conclusion from an article I wrote at TechCrunch. Checkout the full article here.

Enterprise integration has multiple nuances. Integration challenges within an organization, where all systems are controlled by one entity and participants have some degree of trust to each other, are mostly addressed by modern ESBs, BPMs and Microservices architectures. But when it comes to multi-party B2B integration, there are additional challenges. These systems are controlled by multiple organizations, have no visibility of the business processes and do not trust each other. In these scenarios, we see organizations experimenting with a new breed of blockchain-based technology that relies not only on sharing of the protocols and contracts but sharing of the end-to-end business processes and state.

Integration evolution stages

And this trend is aligned with the general direction integration has been evolving over the years: from sharing the very minimum protocols, to sharing and exposing more and more in the form of contracts, APIs and now business processes. This shared integration infrastructure enables new transparent integration models where the previously private business processes are now jointly owned, agreed, built, maintained and standardized using the open-source collaboration model. This can motivate organizations to share business processes and form networks to benefit further from joint innovation, standardization and deeper integration in general.

Which Camel DSL to Choose and Why?

(This post was originally published on Red Hat Developers, the community to learn, code, and share faster. To read the original post, click here.)

Apache Camel is a powerful integration library that provides mainly three things: lot’s of integration connectors + implementation of multiple integration patterns + a higher level Domain Specific Language abstraction to glue all together nicely. While the connectors and pattern choices are use case and feature driven and easy to make, choosing which Camel DSL to use might be a little hard to reason about. I hope this article will be able to guide you in you first Camel journey.

I work for Red Hat Consulting as an Integration architect and one of my primary goals is to help customers get the design and the architecture of their future systems as right as possible, and consequently get the best value out of Apache Camel. One of the common questions I get at the start of every new Camel based project is: “Which Camel DSL should we use? What are the pros and cons of each?”

I have one good news and one more good news for you. First, by choosing to use Apache Camel you have already done the right choice and Camel will turn out to be a very useful toolkit in your arsenal of libraries for lot’s of future use cases and projects to come. And second, the DSL is just a technicality and it will not impact the success of your project and you can always change your mind later and even mix and match.

If you are a part of large company, with multiple independent two-pizza size teams that use Java language here and there, the chances are that some teams are already using Camel. Even in small companies, teams use Camel without being aware of each other as it is useful for all kind of tasks and a small enough library you to add to your pom.xml and use it without a permission from the technical design board. If that is the case, just talk to your colleagues and learn from first hand their experience with their DSL of choice.

If you need a more comprehensive comparison and a reason to choose one of the DSLs, below is a brain dump from multiple engineers developing developing Apache Camel and consultants using Camel at multiple customer projects across the globe. Pick the arguments that are valid in your context and make your choice.

Comparing Apache Camel's XML and Java DSLs

If this table doesn’t give you the straight answer you were looking for, probably the answer is: it doesn’t matter. Camel has multiple DSLs, but there are good reasons for both Java and XML based DSLs to be equally popular. The more important takeaway from here is for developers to get used to think in terms of Pipes and Filters, learn the Enterprise Integration Patterns and their notations. Then using one of the Camel DSLs to express these patterns is a technicality without a technical consequence. Usually it is a team preference and culture based choice, such as “We are a hard-core Java shop and we hate XML” or “Can we do all through drag and drop?”.

All that said, the only advice I can suggest is strive for consistency. Avoid using different DSLs in the same service, even for different services in the same project. And if you can convince everybody in the company to use the same DSL even better.

Check out my Camel Design Patterns book for more Camel related topoics and follow me @bibryam for future blog posts.

Hexagonal Architecture as a Natural fit for Apache Camel

(This post was originally published on Red Hat Developers, the community to learn, code, and share faster. To read the original post, click here.)

There are architectures and patterns that look cool on a paper, and there are ones that are good in practice. Implementing the hexagonal architecture with Camel is both: cool to talk about, and a natural implementation outcome. I love going hexagonal with Camel because it is one of these combinations where the architecture and the tool come together naturally and many end up doing it without realising it. Let’s see why that is the case.

Why go Hexagonal?

Hexagonal architecture is originally described by Alistair Cockburn as an approach for dividing an application into inside and outside parts. Its intent is to move focus from multiple conceptual layers of an application to a distinction between the inside and outside parts of the application. The inside part represents the domain layer or the business logic, and the outside part consists of all the possible incoming or outgoing interaction points of the application. The same architecture is also known as Ports and Adapters as the the connection between the inside and the outside of the application is realized through ports and adapters. The word “port” is inspired by the operating systems ports where any application that conform to the protocol of a port can send or receive signals from an application. In a sense, a port represent a purposeful conversation. And the adapters represent the technology specific implementations of a port. Depending on the business benefits offered through the port, there might be multiple adapters that would like to expose the port using different technologies.

Hexagonal architecture visualized with Enterprise Integration Patterns

Notice that all ports and adapters are fundamentally similar at the architectural level, but Alistair acknowledges that the ports and adapters come up in two flavours: primary and secondary or driving and driven. For example, if there is a simple REST based service that reads and writes to a database, the REST side of the service would be the primary actor port and adapter as it initiates and drives the interactions. The port and adapter for writing to the database side would be the secondary and driven actor as it is not initiating any calls (assuming we are not using any data change capture listeners in which case this adapter would also be a primary one).
Briefly said, hexagonal architecture helps us avoid multi-layered architectures that are prone to end up being baklava architecture (anti-pattern). Instead it pushes us towards simplified separation of concerns, and onion-architecture, clean architecture, and similar.

Why is Camel Hexagonal in Nature?

Let’s look at the two extremes: a layered architecture manages the complexity of a large application by decomposing it and structuring into groups of subtasks of particular abstraction level called layer. Each layer has a specific role and responsibility within the application and changes made in one layer of the architecture usually don’t impact the components of other layers. In practice, this architecture splits an application into horizontal layers, and it is a very common approach for large monolithic web or ESB applications of the JEE world.

Layered architecture compared to Pipes and Filters pattern

On the other extreme is Camel with its expressive DSL and route/flow abstractions. Based on Pipes and Filters pattern, Camel would divide a large processing task into a sequence of smaller independent processing steps (Filters) connected by channel (Pipes). There is no notion of layers that depend on each other, and in fact, because of its powerful DSL, a simple integration can be done in few lines and a single layer only. In practice, Camel routes split your application by use case and business flow into vertical flows rather than horizontal layers. And a typical Camel application is composed of multiple independently working Camel routes that collaborate for achieving the common business goals. As mentioned previously, when working with Camel, services created with it tend to end up as a single layer. Whereas this is fine for most of the simpler cases, applying the hexagonal architecture principles will help with creating better applications when working on large scale projects. What I mean by that is, split your Camel routes into two layers that represent the inside and the outside of the application. The inside of the application is represented by Camel routes that implement the business logic of your integration, and intended to be reused by multiple other routes and protocols. The outside of the application would be implemented by Camel routes that are the adapters in the hexagonal architecture i.e. routes that provide technology specific logic e.g. handle a specific protocol, error handling logic specific for the endpoint, transactionality and recovery actions specific for the endpoint as well.

How to Map Hexagonal Architecture to Camel?

Identify the inside of your application

Even the simplest services created with Camel have some kind of business logic. Usually that is a combination of transforming data, content-based routing, filtering, splitting, aggregating, etc. Very often, none of the out of the box enterprise integration patterns will be applicable and you will have to use your own custom Java bean as part of a Camel route. The awesome part is that Camel is completely non-intrusive and you can develop, test and use Java beans in Camel routes with absolute no dependency on the Camel APIs. Camel bean component will make sure that the bean method parameters are populated with the correct values and also take the return value and put it back into Camel routes. If you have identified the routes containing the elements mentioned above, typically these represent the inside of your application. These kind of routes should not contain logic that is technology and protocol specific. For example avoid using data that is directly populated by components, such as http headers, jms headers, and also error handling retry logic common for the HTTP protocol, compensating action logic, etc. Instead keep this inside Camel routes focused on the business logic only and isolated from outside Camel routes.

Isolate the inside from the outside

In the hexagonal architecture, the inside of the application is reached through ports that abstract conversations. Camel direct component is the perfect implementation of a port. It provides synchronous invocation, the same as a method call in Java. It is not technology and protocol specific, there is no specific data format or schema validation requirement and can be used to pass in and out any kind of data. Typically the preferred data format to pass is a POJO as it is the easiest and most flexible structure to manipulate in a Camel route. But if in your domain, the primary data format is XML, JSON, or anything else, you can keep to such a format as well. No strong rules to follow, but whatever works for you. The only thing that is fixed with direct component is that it is a synchronous interaction model and I think that is the correct one by default. If asynchronicity is required, rather than using SEDA component for a port, it would be better to implement asynchronous logic either as part of the outside route (if asynchronicity is required by an adapter) or in the inside route if it is part of the business logic. But don’t limit the port to asynchronous model only. A port represents a meaningful conversation in the context of a service. That in Camel is represented by direct component which is identified uniquely as a String value in the context of a JVM. One would think that direct component was implemented as a response for Alistairs port definition.

Keep outside out

So we have the business logic of our application implemented as Camel routes accessible only through direct component endpoints as ports. Such a setup allows testing, reusing, and exposing the business logic over multiple protocols to the outside world using other routes. The outside of the application contains any logic that is dependent on the endpoints. Nowadays most common of these are the messaging or file based for asynchronous interaction, and HTTP based for synchronous interaction. But it can also be any of the other over 200 connectors that are present in Camel. Keep in mind that the components you use for the outside, are not only dictating the interaction model, but also usually define the data format, the transaction semantics, the error handling logic, and potentially even other aspects of the applications. For example a SOAP endpoint will perform a schema validation, but consuming a JMS message will require an additional validation step. A transactional endpoint will perform rollback in the case of a failure, but a non-transactional endpoint will require a recovery action, an idempotent endpoint will allow retry, and a non-idempotent endpoint will not. I have described these kind of considerations and other related Camel use cases in more details in the Camel Design Patterns book. Putting it all together, a Camel based service that exposes some business functionality over SOAP and JMS is visualized below. The same business functionality is accessible through direct component for JMS and SOAP based routes. Also, on the right hand side, the same route is using an email notifications port and adapter for sending emails.

Ports and Adapters based Camel service

 Notice that, outside routes are not only on the consumer side, they are also on the producer side, i.e. routes that send messages to other systems (remember driving and driven ports/adapters). The intent of the outside routes is to represent the various adapters that should handle everything that is outside specific: protocol, data format, additional logic specific for the endpoint. In addition, an outside route should prepare the data in format that is expected by the port by populating expected headers and the message body. This would allow the same port to be reused by multiple adapter routes. This includes also test fixtures in Camel for unit testing Camel routes, and even the error handling code. The error handling constructs in Camel (no doTry, doCatch, doFinally, but the onException construct) is actually representing a port that is automatically called by the framework on different types of exceptional conditions. Such a concept doesn’t exist in the Java language, but in Camel it is a very commonly used execution path for unhappy scenarios. And treating the error handling flow as just another port in your application (even if it is not called by you but the framework on certain occasion), will help you to reuse it for common error handling across multiple Camel routes.

In Summary

There are no clear rules or guidelines on how to compose an application with Camel routes. Defining those at design time usually limits the creativity of developers at implementation time, and not having guidelines can be a recipe for a spaghetti architecture. In this line of thought, I think hexagonal architecture is sufficiently lightweight, doesn’t kill creativity and imagination during implementation by forcing specific structure. At the same time, it provides just enough guidance for structuring routes. And the best part is, it naturally fits the Camel programing model.
My suggestion would be start with VETRO pattern (Validate, Enrich, Transform, Route, Operate), then apply hexagonal architecture style (Edge Component Pattern as described in Camel Design Patterns book). This is a good starting point for structuring Camel routes for the happy paths. Then pay special attention to achieving data consistency with the various error handling and recovery patterns. And don’t forget, there are no best practices, but only good practices in a context. Focus on your context and Camel will be on your side.
Follow me @bibryam for future blog posts on related topics.

Short Retry vs Long Retry in Apache Camel

(This post was originally published on Red Hat Developers, the community to learn, code, and share faster. To read the original post, click here.)

Camel Design Patterns book describes 20 patterns and numerous tips and best practices for designing Apache Camel based integration solutions. Each pattern is based on a real world use case and provides Camel specific implementation details and best practises. To get a feel of the book, below is an extract from the Retry Pattern from the book describing how to do Short and Long retires in Apache Camel.

Context and Problem

By their very nature integration applications have to interact with other systems over the network. With dynamic cloud-based environments becoming the norm, and the microservices architectural style partitioning applications into more granular services, the successful service communication has become a fundamental prerequisite for many distributed applications. Services that communicate with other services must be able to handle transient failures that can occur in downstream systems transparently, and continue operating without any disruption. As a transient failure can be considered an infrastructure-level fault, a loss of network connectivity, timeouts and throttling applied by busy services, etc. These conditions occur infrequently and they are typically self- correcting, and usually retrying an operation succeeds.

Forces and Solution       

Reproducing and explaining transient failures can be a difficult task as these might be caused by a combination of factors happening irregularly and related to external systems. Tools such as Chaos Monkey can be used to simulate unpredictable system outages and let you test the application resiliency if needed. A good strategy for dealing with transient failures is to retry the operation and hope that it will succeed (if the error is truly transient, it will succeed; just keep calm and keep retrying).
To implement a “retry” logic there are a few areas to consider:            

Which failures to retry?

Certain service operations, such as HTTP calls and relational database interactions, are potential candidates for a retry logic, but further analysis is needed before implementing it. A relational database may reject a connection attempt because it is throttling against excessive resource usage, or reject an SQL insert operation because of concurrent modification. Retrying in these situations could be successful. But if an relational database rejects a connection because of wrong credentials, or an SQL insert operation has failed because of foreign key constraints, retrying the operation will not help. Similarly with HTTP calls, retrying a connection timeout or response timeout may help, but retrying a SOAP Fault caused by a business error does not make any sense. So choose your retries carefully.

How often to retry?

Once a retry necessity has been identified, the specific retry policy should be tuned to satisfy the nature of both applications: the service consumer with the retry logic and the service provider with the transient failure. For example, if a real time integration service fails to process a request, it might be allowed to do only few retry attempts with short delays before returning a response, whereas a batch-based asynchronous service may be able to afford to do more retries with longer delays and exponential back off. The retry strategy should also consider other factors such as the service consumption contracts and the SLAs of the service provider. For example, a very aggressive retry strategy may cause further throttling and even a blacklisting of a service consumer, or it can fully overload and degrade a busy service and prevent it from recovering at all. Some APIs may give you an indication of the remaining request count for a time period and blacklisting information in the response, but some may not. So a retry strategy defines how often to retry and for how long before you should accept the fact that it is a non-transient failure and give up.

Idempotency

When retrying an operation, consider the possible side effects on that operation. A service operation that will be consumed with retry logic should be designed and implemented as idempotent. Retrying the same operation with the same data input should not have any side effects. Imagine a request that has processed successfully, but the response has not reached back. The service consumer may assume that the request has failed and retry the same operation which may have some unexpected side effects.

Monitoring

Tracking and reporting retries is important too. If certain operations are constantly retried before succeeding or they are retried too many times before failing, these have to be identified and fixed. Since retries in a service are supposed to be transparent to the service consumer, without proper monitoring in place, they may remain undetected and affect the stability and the performance of the whole system in a negative way.

Timeouts and SLAs

When transient failures happen in the downstream systems and the retry logic kicks in, the overall processing time of the retrying service will increase significantly. Rather than thinking about the retry parameters from the perspective of the number of retries and delays, it is important to drive these values from the perspective of service SLAs and service consumer timeouts. So take the maximum amount of time allowed to handle the request, and determine the maximum number of retries and delays (including the processing time) that can be squeezed into that time frame.

Mechanics

There are a few different ways of performing retries with Camel and ActiveMQ.

Camel RedeliveryPolicy (Short Retry)

This is the most popular and generic way of doing retries in a Camel. A redelivery policy defines the retry rules (such as the number of retries and delays, whether to use collision avoidance and an exponential backoff multiplier, and logging) which can then be applied to multiple errorHandler and onException blocks of the processing flow. Whenever an exception is thrown up, the rules in the redelivery policy will be applied.

Camel RedeliveryPolicy example

The key differentiator of the retry mechanism is that Camel error handling logic will not retry the whole route, but it will retry only the failed endpoint in the processing flow. This is achieved thanks to the channels that connect the endpoints in the Camel route. Whenever an exception is thrown up by the processing node, it is propagated back and caught by the channel, which can then apply various error handling policies. Another important difference here is that Camel-based error handling and redelivery logic is in-memory, and it blocks a thread during retries, which has consequences. You may run out of threads if all threads are blocked and waiting to do retries. The owner of the threads may be the consumer, or some parallel processing construct with a thread pool from the route (such as a parallel splitter, recipient list, or Threads DSL). If, for example, we have an HTTP consumer with ten request processing threads, a database that is busy and rejects connections, and a RedeliveryPolicy with exponential backoff, after ten requests all the threads will end up waiting to do retries and no thread will be available to handle new requests. A solution for this blocking of threads problem is opting for asyncDelayedRedelivery where Camel will use a thread pool and schedule the redelivery asynchronously. But the thread pool stores the redelivery requests in an internal queue, so this option can consume all of the heap very quickly. Also keep in mind that there is one thread pool for all error handlers and redeliveries for a CamelContext, so unless you configure a specific thread pool for long-lasting redelivery, the pool can be exhausted in one route and block threads in another. Another implication is that because of the in-memory nature of the retry logic, restarting the application will lose the retry state, and there will be no way of distributing or persisting this state.
Overall, this Camel retry mechanism is good for short-lived local retries, and to overcome network glitches or short locks on resources. For longer-lasting delays, it is a better option to redesign the application with persistent redeliveries that are clustered and non-thread-blocking (such a solution is described below).

ActiveMQ Broker Redelivery (Long Retry)

This retry mechanism has different characteristics to the previous two since it is managed by the broker itself (rather than the message consumer or the Camel routing engine). ActiveMQ has the ability to deliver messages with delays thanks to its scheduler. This functionality is the base for the broker redelivery plug-in. The redelivery plug-in can intercept dead letter processing and reschedule the failing messages for redelivery. Rather than being delivered to a DLQ, a failing message is scheduled to go to the tail of the original queue and redelivered to a message consumer. This is useful when the total message order is not important and when throughput and load distribution among consumers is.

ActiveMQ redelivery example

The difference to the previous approaches is that the message is persistent in the broker message store and it would survive broker or Camel route restart without affecting the redelivery timings. Another advantage is that there is no thread blocked for each retried message. Since the message is returned back to the broker, the Competing Consumers Pattern can be used to deliver the message to a different consumer. But the side effect is that the message order is lost as the message will be put at the tail of the message queue. Also, running the broker with a scheduler has some performance impact. This retry mechanism is useful for long-delayed retries where you cannot afford to have a blocked thread for every failing message. It is also useful when you want the message to be persisted and clustered for the redelivery.
Notice that it is easy to implement the broker redelivery logic manually rather than by using the broker redelivery plug-in. All you have to do is catch the exception and send the message with an AMQ_SCHEDULED_DELAY header to an intermediary queue. Once the delay has passed, the message will be consumed and the same operation will be retried. You can reschedule and process the same message multiple times until giving up and putting the message in a backoff or dead letter queue.

Side note - I know, shameless plug, but I'm pretty excited about my book on this topic. You can check it out here at a 40% discount until end of June! And hope you like it.

A geeky laptop skin

Recently we had a company meetup where the team leads talked about the Red Hat culture, mission and values. Inspired by all the talks and the stickers on the table that day, I've created my first laptop skin. And the result was this:

Red Hat - the Open Source Catalyst

In case you want to create something similar here the steps I took:

1. Pick some nerdy words. I picked all the middleware project names that Red Hat contributes to (I've never realized there were that many).  Removed some that are not so cool any longer (EJB3) and added few other hot projects (Kubernetes, Apache isis, OFBiz).

2. Created the cloud tag using the above words and the shadowman picture at Tagul. You can log in and edit my existing desing for a quicker sticker.

3. Exported the image from step 2 and created a Mac skin at wrappz which costs around £15. (The following discount code should give you 20% off Bilgin20 and some royalties for me).

Attached is the transparent image used for the skin and the editable version is at Tagul.

From Fragile to Antifragile Software

(This post was originally published on Red Hat Developers, the community to learn, code, and share faster. To read the original post, click here.)

One of my favourite books is Antifragile by Nassim Taleb where the author talks about things that gain from disorder. Nacim introduces the concept of antifragility which is similar to hormesis in biology or creative destruction (accessible version creative-destruction) in economics and analyses it characteristics in great details. If you find this topic interesting, there are also other authors who have examined the same phenomenon in different industries such as Gary Hamel, C. S. Holling, Jan Husdal. The concept of antifragile is the opposite of the fragile. A fragile thing such as a package of wine glasses is easily broken when dropped but an antifragile object would benefit from such stress. So rather than marking such a box with "Handle with Care", it would be labelled "Please Mishandle" and the wine would get better with each drop (would be awesome woulnd't it).

It didn't take long for the concept of antifragility to be used also for describing some of the software development principles and architectural styles. Some would say that SOLID prinsiples are antifragile, some would say that microservices are antifragile, and some would say software systems cannot be antifragile ever. This article is my take on the subject.

According to Taleb, fragility, robustness, resilience and antifragility are all very different. Fragility involves loss and penalisation from disorder. Robustness is enduring to stress with no harm nor gain. Resilience involves adapting to stress and staying the same. And antifragility involves gain and benefit from disorder. If we try to relate these concepts and their characteristics to software systems, one way to define them would be as the following.

Disorder 

Different systems are affected by different kind of disorder, such as stress, time, change, volatility, debt, etc. For software systems, the main disorder is the change. The business is in a constant changing environment and the software needs to adapt to the business needs quickly. That is implementing new requirements, changes to existing functionality, even creating new business opportunities through innovation. A software system has to change all the time, otherwise it is obsolete.
Apart from development time challenges, there are also runtime challenges for software systems too. Software systems are created and exists to add value by running in a production environment. And while doing so, they are under stress by end users and other systems. This is another kind of disorder, that software systems have to deal with.

Fragile 

This property describes systems that suffer when put under stress. Imagine a software project that is not easy to change at development time. For example if it is not easy to extend, modify and deploy to the production environment. Or a system that is not able to handle unexpected user inputs or external system failures and breaks easily. That's is a fragile system that is harmed by stress and penalised by change, a good example of fragile.

Robust

This is a system that can continue functioning in the presence of internal and external challenges without adaptation. Every system is robust up to a level. For example a bottle is robust until it reaches the level of breaking point. A software system can be made robust to handle unanticipated user input, or failures in external systems. For example handling NPE in Java, using try-catch-finally statements to handle unreliable invocations, having a thread pool to handle concurrent users, creating network connections using timeouts, are all examples of robustness for a software system. But a robust system doesn't adapt to a chaining environment and when the stress and change threshold is reached it would break. The qualities that define the robustness vary from system to system. An ATM for example needs to be robust and not fail in the middle of a transaction, whereas, a media streaming service can drop a frame or two, as long as it continues streaming under stress.

Resilient

A system is resilient when it can adapt to internal and external challenges by changing its method of operation. The key here is that the system is responding to stress by changing its internal behaviour rather than resisting stress with a predefined buffer.

• A typical example here is the circuit breaker pattern which changes its internal state to adapt to the external system behaviour to protect itself. 
• Another example would be using a retry mechanism with some backoff algorithm to handle transient failures in external systems.
• A different technique for creating resilient systems is through graceful degradation, both on the UI and the server side. Rendering UI based on the user agent capabilities, or failing fast on the server side and fallback to some default values are commonly used techniques to adapt to failures.
• Systems with self healing and autorepair capabilities are another example of resiliency. These systems are self-aware and can detect abnormalities and take corrective actions. For example Kubernetes/OpenShift will perform regular liveness checks for the running Docker containers and if they detect any anomalies they will restart the container and perform necessary backoffs until the application stabilises. This is another mechanism to cope with stress and improve application resilience.

Overview

Before looking at the next level of software evolution - the antifragile systems, let's visualise and summarise different kind of software system characteristics.

• A fragile system is difficult to modify, and cannot cope with a changing environment. Even if it provides some value when used in a stable non changing environment, when faced with further stress and change, it quickly turns into a liability. Many organisations have applications (mainframes for example) which are impossible to change, very expensive to maintain, but still running on high cost as they are very critical to the business.
• A robust system is the one that is implemented with certain buffers to handle change and stress. So when the stress level increases, it can withstand it for up to a level without losing its capabilities and still provide a good value. But a robust system does not adapt, and if the stress and change levels continue raising, such a system can also stops providing benefit and may turn into liability and run on loss.
• A resilient system can handle more stress and change as it is designed and implemented with stress in mind and adaptability features. Even if it is not benefiting from stress, it can survive lot's of different kind of stress and change and provide value up to a greater degree.
• An antifragile system is created with change in mind and it feeds from stress and change. It is much harder to create such a system (it is not a software system but a social-technical system) but once it is in place, it drives the business based on change, and even creates the change.

Antifragile

Many things in life are antifragile, such as the human body. When stressed at the right level, a muscle or bone wold come back stronger. But can a software system be antifragile? Certainly there are some tools, platforms, architectural styles, methodologies that can help create software with antifragile characteristics. Let's see some of the more popular ones.

Auto scaling feature allows applications to handle increasing load by creating more instances of the application. To achieve that, the software system have to able to measure and then react to change and stress. Some good examples here are AWS autoscaling of EC2 instances at infrastructure level, and OpenShift autoscaling of application containers for the application level. This is a feature that transitions applications from resiliency to antifragility since the software system is shifting resources from one part of the system into another to respond to stress.

Microservices. According to Taleb, at times of stress, the large is doomed to breaking. And that phenomen has been observed with mammals, corporations, administrations, etc. In software and large projects, this behaviour has been observed even more often. The bigger a software project is, the harder it becomes to change and react to stress. Microservices is an architecture styles that allows easier change by having autonomous services with well defined APIs - features that allow change. Russ Miles is a strong believer and proponent of Antifragile Software through Microservices (here is an intro video from him).

Chaos engineering is a technique to create antifragility by evolving systems to survive chaos. Rather than waiting for things to break at the worst possible time, the idea of chaos engineering is to proactively inject failures in order to be prepared when disaster strikes. Netflix’s Simian Army is a very good materialization of this technique, designed to generate failures and help isolate system’s weaknesses.

Continuous deployments to a production environment creates continuous partial system failures and forces organisations to react better to failures through redundancy, rolling upgrades, rollbacks, and avoiding single points of failure. Other techniques such as canary release, blue-green deployments, are used to reduce the risk of introducing new software into production environment. Some other methods such as A/B testing even allows experimenting with change and measuring its effect, in order to gain from change.

The human element

Antifragility is not an universal characteristics. Different systems are antigragile towards different kind of disorder. For example Chaos Monkey will make your system antifragile towards EC2 deaths, and autoscaller will make your system respond to specific type of load. But your systems will not be antifragile towards other kinds of stress. And if you look above at the different ways of introdusing antifragility into software systems, all of them are means for making the social-technical system antifragile (and not only the software system):

• Chaos engineering forces human feedback to the injected randomness and makes the system antifragile.
• Microservices alone do not make a software system antifragile. But microservices combined with appropriate organisational and team structure enables antifragility. If microservices are a way for architecting applications into autonomous structures, DevOps is a means for organizing teams into similar structures. You need both in order to benefit from them and gain from disorder.
• Continuous deployment pipeline is a tool that allows teams to react to stress faster, introduce or retract change faster and generally use the stress as the driver.
• Similarly, iterative development is not enough to benefit from changing environment. But iterative development with open and honest retrospective rituals is.

Innovate

If it takes few weeks to create a new developer environment, you can not react to change. If it takes three months to release a new feature, you can not react to change. If your ops team is watching the metrics dashboard and manually scaling applications up and down, you cannot embrace change. If the team is hardly catching up with the change, there is no way to gain from change. But once you put an appropriate organisational structure, the right tools and culture in place, then you can start gaining from change. Then you can afford having Friday Hackathons, then you can start exploring open source projects and start contributing to them, then you can start open sourcing your internal projects and benefit from a community, and generally be the change itself. And why not the Netflix or the Amazon of tomorrow.

Visualizing Integration Applications

(This post was originally published on Red Hat Middleware Blog. To read the original post, click here.)
Since I've changed role and started performing architect duties, I have to draw more boxes and arrows than write code. There are ways to fight that, like contributing to open source projects during sleepless nights, POCs, demos, but drawing boxes to express architectures and designs is still big part of it. This post is about visualising distributed messaging/SOA/microservices applications in agile (this term has lost its meaning, but cannot find a better one in this case) environments. What I like about the software industry in recent years is that the majority of organisations I've worked with, value the principles behind lean and agile software development methodologies. As long as it is practical, everyone strives to deliver working software (rather than documentation), deliver fast (rather than plan for a long time), eliminate waste, respond to change, etc. And there are management practises such as Scrum and Kanban, and Technical Practises from Extreme programming (XP) methodology such as unit testing, pair programing, and other practises such as CI, CD, DevOps to help implement the aforementioned principles. In this line of thinking, I decided to put together a summary of the design tools and diagrams I find useful in my day to day job while working with distributed Systems.

Issues with 4+1 View Model and death by UML

Every project kicks off with big ambitions, but there is never enough time to do things perfectly, and at the end we have to deliver whatever works. And that is a good thing, it is the way the environment helps us avoid gold plating, YAGNI, KISS, etc. so we do just enough and adapt to chance.

Looking back, I can say that most of the diagrams I've seen around are inspired by 4+1 view model of Philippe Kruchten which has Logical, Development, Process and Physical views.

4+1 View Model

I quite like the ideas and the motivation behind this framework: using separate views and perspectives to address specific set of constraints and targeting the different stakeholders. That is a great way of describing complex software architectures. But I have two issues with using this model for integration applications.

Diagram applicability

Typically these views are expressed through UML, and for each view, you have to use one or more UML diagrams. The fact that I have to use 15 types of UML diagrams to communicate and express a system architecture in an accessible way, defeats its purpose.

Death by UML

With such a complexity, the chances are that there are only one or two people in the whole organisation who has the tools to create, ability to understand and maintain these diagrams. And having hard to interpret, out of date diagrams is as useful as having out of date gibberish documentation. These diagrams are too complex and with limited value, and very quickly they turns into liability that you have to maintain rather than asset expressing the state of a constantly changing system.
Another big drawback is that the existing UML diagram types are primarily focused on describing object-oriented architectures rather than Pipes and Filters architectures. The essence of messaging applications is around interaction styles, routing, data flow rather than structure. Class, object, component, package, and other diagrams are of less value for describing a Pipes and Filters based processing flows. Behavioural UML diagrams such as activity and sequence get closer, but still cannot express easily concepts such filtering and content based routing which are fundamental part of integration applications.

View applicability 

Having different set of views for a system, to address different concerns is a great way of expressing intend. But the existing views of 4+1 model doesn't reflect the way we develop and deploy software nowadays. The idea that you have a logical view first, which then leads to development and process view, and those lead to physical view is not always the case. The systems development life cycle, is not following the (waterfall) sequence of requirement gathering, designing, implementing and maintaining.

Software Development Lifecycle

Instead other development methodologies such as agile, prototyping, synchronise and stabilise, spike and stabilise are used too. In addition to the process, the stakeholders are changing too. With practises such as DevOps, developers have to know about the final physical deployment model, operations have to know about the application processing flows too. Modern architectures such as microservices affect the views too. Knowing one microservice is in a plethora of a services is not very useful. Knowing too much about all the services is not practical either. Having the right abstraction level to have a system wide view with just enough details becomes vital.

Practical Visualisation for Integration Applications

There are some good tips for drawing in general. The closest thing that has been working for me is described by Simon Brown as C4 model. (You should also get a free copy of Simon's awesome The Art of Visualising Software Architecture book). In his model, Simon is talking about the importance of a common set of abstractions rather than common notation (such as UML) and then using simple set of diagrams for different level of abstractions: system context diagram, container diagram, component diagram and class diagram. I quite like this "Outside-In" approach, where you first have 10000 foot view and with each next level, going deeper with more detailed views.
C4 is also not an exact match for middleware/integration applications either, but it is getting closer. If we were to use C4 model, then system context diagram would be one box that says ESB (or middleware, MOM, microservices) with tens of arrows from north to south. Not very useful. Container diagram is quite close, but the term container is so overloaded (VM, application container, docker container) which makes it less useful for communication. Component and class diagrams are also not a good fit as Pipes and Filter architectures are focused around Enterprise Integration Patterns, rather than classes and packages.
So at the end, what is it that worked for me? It is the following 3 types of diagrams which abbreviate as SSD (not as cool as C4):  System Context Diagram, Service Design Diagram and Deployment Diagram.

System Context Diagram

The aim of this model is to show all the services (whether they are SOA, Microservices) with their inputs and outputs. Ideally having the external systems on the north, the services in the middle section, and internal services in the south. Or you could use both external and internal services on both side of the middleware layer as shown below. Also having the protocol (such as HTTP, JMS, file) on the arrows, with the data format (XML, JSON, CSV) gives useful context too, but it is not mandatory. If there are too many services, you can leave the protocol and the data format for the service level diagrams. I use the direction of the arrow to indicate which service is initiating the call rather than the data flow direction.

System Context Diagram

Having such a diagram gives a good overview of the scope of a distributed system. We can see all the services, the internal and external dependencies, the types of interaction (with protocol and data format), and the call initiator.

Service Design Diagram

The aim of this diagram is to show what is going on in each box representing a middleware service from the System Context Diagram. And the best diagram for this is to use EIP icons and connect those as message flows. A service may have a number of flows, support a number of protocols, implement real time and/or batch behaviour.

Service Design Diagram

At this level, we want to show all possible data flows implemented by a specific service, from any source to any destination.

Deployment Diagram

The previous two diagrams are the logical views of the system as a whole and each service separately. With the deployment diagram, we want to show where each service is going to be deployed. May be there will be multiple instances of the same service running on multiple hosts. May be some services will be active on one host, and passive on the other. May be there will be a load balancer fronting the services, etc.

Deployment Diagram

The deployment diagram is supposed to show how individual services and the system as a whole relates to the host systems (regardless whether that is physical or virtual).

What tools do I use?

The System Context and the Deployment Diagrams are composed only of boxes and arrows and do not require any specail tools. For the Service Design Diagram, you will need a tool that has the Enterprise Integration Pattern icons installed. So far, I have seen the following tools with EIP icon support:

• Mac OS: OmniGraffle with icons from graffletopia.This is what I've used to create all the diagrams for Camel Design Patterns book.
• Windows: Enterprise Architect by Sparx Systems with EIP icons by Harald Westphal. There are also MS Visio Stencils here.
• Web: LucidCharts which ships EIP icons by default. That is the easiest to use tool and accessible from everywhere. It is my favourite tool (with MS Visio import/export options for Windows users), and has free account plans to start with.
• Web: DrawIO another web tool with EIP icons. The beauty of this tool is that it forces you to use your own storage options for the diagrams, such as: google drive, dropbox, locally, etc. So you own the diagrams and keep them safe.
• Canva: not specifically for EIP diagrams, but in general this site has a nice templates and photos to use in various presentations. Check it out.

Other development tools that could be also used for creating EIP diagrams are:

• Red Hat JBoss Developer Studio
• EIP Designer by Sirius
• Talend Open Studio for ESB
• Hawtio - the Camel plugin can visualize running Camel routes as EIP diagrams.
  

System Context Diagram is useful to show the system wide scope and reach of the services, Service Design Diagram is good for describing what a service does, and Deployment Diagram is useful mapping all that into something physical. In IT, we can expand work and fill up all the available time with things to do. I'm sure given more time, we can invent 10 more useful views. But w/o the above three, I cannot imagine myself describing an integration application. As Antoine de Saint-Exupery put it long ago: "Perfection is finally attained not when there is no longer anything to add but when there is no longer anything to take away."

Create Resilient Camel applications with Hystrix DSL

(This post was originally published on Red Hat Developers, the community to learn, code, and share faster. To read the original post, click here.)
 
Apache Camel is a mature integration library (over 9 years old now) that implements all the patterns from Enterprise Integration Patterns book. But Camel is not only an EIP implementation library, it is a modern framework that constantly evolves, adds new patterns and adapts to the changes in the industry. Apart from tens of connectors added in each release, Camel also goes hand-in-hand with the new features provided by the new versions of Java language itself and other Java frameworks. With time some architectural styles such as SOA and ESB lose attraction and new architectural styles such as REST, Microservices get popular. To enable developers do integrations using these new trends, Camel responds by adding new DSLs such the REST DSL and new patterns such as the Circuit Breaker, and components such as Spring Boot. And that's not all and we are nowhere near done. With technologies such as Docker containers and Kubernetes, the IT industry is moving forward even faster now, and  Camel is evolving too in order to ease the developers as it always has been. To get an idea of kinda tools you would need to develop and run applications on Docker and Kubernetes, check out the Fabric8 project and specifically tools such as the Docker Maven plugin, Kubernetes CDI extension, Kubernetes Java client, Arquilian tests for Kubernetes, etc. Exciting times ahead with lot's of cool technologies, so let's have a look at one of those: Camel and the Hystrix based circuit breaker.

Two Circuit Breakers in Camel, which one to chose?

Two years ago, when I first read Release It from Michael Nygard, I couldn't stop myself implementing the Circuit Breaker pattern in Camel. Usually I drive my contributions by real customer needs, but Circuit Breaker pattern is so elegant, I had to do it. To implement it in a non-intrusive manner, I have added it as a Camel Load Balancer strategy. Here how simple it is:
The DSL above is self describing: if the number of MyCustomExceptions thrown by mock:result endpoint reaches the threshold number, the CircuitBreaker goes to open state and starts rejecting all requests. After 1000ms it moves to halfOpenAfter state and the result of the first request in this state will define its next state as closed or open. It is the simplest possible implementation of the CircuitBreker you can imagine, but still useful.

Since then, the Microseservices architecture has became more popular, and so is the Circuit Breaker Pattern and its java implementation Hystrix. At some point Raúl Kripalani started the Hystrix implementation in Camel and put all the ground work in place, but with time it lost momentum. Then seeing the same request again and again from different customers, I took the relay and continued the work and pushed to Camel a Hystrix component. Seeing the feedback from the community, it still didn't feel as elegant as it could be. Then Claus stepped in and made Hystrix part of the Camel DSL by turning it into an EIP (rather than component). So how does it look like to create a Hystrix based Circuit Breaker in Camel now?

In the example above, you can see only very few of the available options for a Circuit Breaker, to see all of them checkout the offical documents and try out the example application Claus put in place.

Based on this example you may think that Hystrix is part of Camel core, but it is not. Camel core is still this light and without dependencies to third party libraries. If you want to use the Hystrix based Circuit Breaker, you need to add camel-hystrix dependency to your dependencies as it is with any other non-core component and make it available at runtime.

Fail Fast, Fallback, Bulkhead, Timeout and more

The Hystrix library implements more than Circuit Breaker patter. It also does bulkheading, request caching, timeouts, request collapsing, etc. The implementation in Camel does not support request collapsing and caching as these are done using other patterns and components available in Camel already. Request collapsing in Camel can be done using Aggregator EIP and caching can be done using cache components such as Redis, Inifinspan, Hazelcast, etc. The Hystrix DSL in Camel offers around 30 configuration options supported by Hystrix, also exposes metrics over JMX and/or REST for the Hystrix Dashboard.

As a final note, don't foget that to create a true resilient application, you need more than Hystrix. Hystrix will do bulkheading at thread pool level, but that is not enough if you don't apply the same principle at process, host and phisical machine level. To create a create a resilient distributed system, you will need to use also Retry, Throttling, Timeout... and other good bractices some of which I have described in Camel Design Patterns book.

To get some hands on feeling of the new pattern, check the example and then start defending your Camel based Microservices with Hystrix.