Microservices

Assuming that the majority of providers using microservices architecture,

• One can scale the system by increasing the number of instances of service by bringing up more containers.
• One can also apply to cache at microservice layer which can be easy to manage as an invalidation of the cache can be done very easily as the microservice will be the single source of truth.
• Caching can also be introduced at the API Gateway layer where one can define caching rules like when to invalidate the cache.
• One can also shut down some containers when the requirement is less. That is, scale down.

Microservices Architecture is a style of developing a scalable, distributed & highly automated system made up of many small autonomous services. It is not a technology but a new trend evolved out of SOA.

There is no single definition that fully describes the term "microservices". Some of the famous authors have tried to define it in the following way:

1. Microservices are small, autonomous services that work together.
2. Loosely coupled service-oriented architecture with bounded contexts.
3. Microservice architecture is a natural consequence of applying the single responsibility principle at the architectural level.

Microservices are a continuation to SOA.

SOA started gaining ground due to its distributed architecture approach and it emerged to combat the problems of large monolithic applications, around 2006.

Both (SOA and Microservices) of these architectures share one common thing that they both are distributed architecture and both allow high scalability. In both, service components are accessed remotely through remote access protocol (RMI, REST, SOAP, AMQP, JMS, etc.). both are modular and loosely coupled by design and offer high scalability. Microservices started gaining buzz in late 2000 after the emergence of lightweight containers, Docker, Orchestration Frameworks (Kubernetes, Mesos). Microservices differ from SOA in a significant manner conceptually -

1. SOA uses Enterprise Service Bus for communication, while microservices uses REST or some other less elaborate messaging system (AMQP, etc). Also, microservice follow "Smart endpoints and dumb points", which means that when a microservice needs another one as a dependency, it should use it directly without any routing logic/components handling the pipe.
2. In microservices, service deployment and management should be fully automated, whereas SOA services are often implemented in deployment monoliths.
3. Generally, microservices are significantly smaller than what SOA tends to be. Here we are not talking about the codebase here because few languages are more verbose than the other ones. We are talking about the scope (problem domain) of the service itself. Microservices generally do one thing in a better way.
4. Microservices should own their own data while SOA may share a common database. So one Microservices should not allow another Microservices to change/read its data directly.
5. Classic SOA is more platform-driven, while we have a lot of technology independence in case of microservices. Each microservice can have its own technology stack based on its own functional requirements. So microservices offers more choices in all dimensions.
6. Microservices make an as little assumption as possible on the external environment. A Microservice should manage its own functional domain and data model.

Bounded Context is a central pattern in Domain-Driven Design. In Bounded Context, everything related to the domain is visible within context internally but opaque to other bounded contexts. DDD deals with large models by dividing them into different Bounded Contexts and being explicit about their interrelationships.

Monolithic Conceptual Model Problem 

A single conceptual model for the entire organization is very tricky to deal with. The only benefit of such a unified model is that integration is easy across the whole enterprise, but the drawbacks are many, for example:

1. At first, it's very hard to build a single model that works for the entire organization.
2. It's hard for others (teams) to understand it.
3. It's very difficult to change such a shared model to accommodate the new business requirements. The impact of such a change will be widespread across team boundaries.
4. Any large enterprise needs a model that is either very large or abstract.
5. Meaning of a single word may be different in different departments of an organization, so it may be really difficult to come up with a single unified model. Such a model, even if created, will lead to a lot of confusion across the teams.

1. High Cohesion - Small and focussed on doing one thing well. Small does not mean less number of lines of code because few programming languages are more verbose than others, but it means the smallest functional area that a single microservices caters to.
2. Loose Coupling - Autonomous - the ability to deploy different services independently, and reliability, due to the ability for a service to run even if another service is down.
3. Bounded Context - A Microservice serves a bounded context in a domain. It communicates with the rest of the domain by using an interface for that Bounded context.
4. Organisation around business capabilities instead of around technology.
5. Continuous Delivery and Infrastructure automation.
6. Versioning for backward compatibility. Even multiple versions of same microservices can exist in a production environment.
7. Fault Tolerance - if one service fails, it will not affect the rest of the system. For example, if a microservices serving the comments and reviews for e-commerce fails, the rest of the website should run fine.
8. Decentralized data management with each service owning its database rather than a single shared database. Every microservice has the freedom to choose the right type of database appropriate for its business use-case (for example, RDBMS for Order Management, NoSql for catalogue management for an e-commerce website)
9. Eventual Consistency - event-driven asynchronous updates.
10. Security - Every microservice should have the capability to protect its own resources from unauthorized access. This is achieved using stateless security mechanisms like JSON Web Token (JWT pronounced as jot) with OAuth2.

Embracing microservices architecture brings many benefits compared to using monolith architecture in your application, including:

• Autonomous Deployments

The decentralized teams working on individual microservices are mostly independent of each other, so changing a service does not require coordination with other teams. This can lead to significantly faster release cycles. It is very hard to achieve the same thing in a realistic monolithic application where a small change may require regression of the entire system.

• Culture Shift

Microservices style of system architecture emphasizes on the culture of freedom, single responsibility, autonomy of teams, faster release iterations and technology diversification.

• Technology Diversification

Unlike in monolithic applications, microservices are not bound to one technology stack (Java, .Net, Go, Erlang, Python, etc). Each team is free to choose a technology stack that is best suited for its requirements. For example, we are free to choose Java for a microservice, c++ for others and Go for another one.

• DevOps Culture

The term comes from an abbreviated compound of "development" and "operations". It is a culture that emphasizes effective communication and collaboration between product management, software development, and operations team. DevOps culture, if implemented correctly can lead to shorter development cycles and thus faster time to market.

Polyglot persistence is all about using different databases for different business needs within a single distributed system. We already have different database products in the market each for a specific business need, for example:

• RDBMS

Relational databases are used for transactional needs (storing financial data, reporting requirements, etc.)

• MongoDB

Documented oriented databases are used for documents oriented needs (for e.g. Product Catalog). Documents are schema-free so changes in the schema can be accommodated into the application without much headache.

• Cassandra/Amazon DynamoDB

Key-value pair based database (User activity tracking, Analytics, etc.). DynamoDB can store documents as well as key-value pairs.

• Redis

In memory distributed database (user session tracking), its mostly used as a distributed cache among multiple microservices.

• Neo4j

Graph DB (social connections, recommendations, etc)

Benefits of Polyglot Persistence are manifold and can be harvested in both monolithic as well as microservices architecture. Any decent-sized product will have a variety of needs which may not be fulfilled by a single kind of database alone. For example, if there are no transactional needs for a particular microservice, then it's way better to use a key-value pair or document-oriented NoSql rather than using a transactional RDBMS database.

References: https://martinfowler.com/bliki/PolyglotPersistence.html

The Twelve-Factor App is a recent methodology (and/or a manifesto) for writing web applications which run as a service.

• Codebase

One codebase, multiple deploys. This means that we should only have one codebase for different versions of a microservices. Branches are ok, but different repositories are not.

• Dependencies

Explicitly declare and isolate dependencies. The manifesto advises against relying on software or libraries on the host machine. Every dependency should be put into pom.xml or build.gradle file. 

• Config

Store config in the environment. Do never commit your environment-specific configuration (most importantly: password) in the source code repo. Spring Cloud Config provides server and client-side support for externalized configuration in a distributed system. Using Spring Cloud Config Server you have a central place to manage external properties for applications across all environments. 

• Backing services 

Treat backing services as attached resources. A microservice should treat external services equally, regardless of whether you manage them or some other team. For example, never hard code the absolute url for dependent service in your application code, even if the dependent microservice is developed by your own team. For example, instead of hard coding url for another service in your RestTemplate, use Ribbon (with or without Eureka) to define the url: 

• Release & Run 

Strictly separate build and run stages. In other words, you should be able to build or compile the code, then combine that with specific configuration information to create a specific release, then deliberately run that release. It should be impossible to make code changes at runtime, for e.g. changing the class files in tomcat directly. There should always be a unique id for each version of release, mostly a timestamp. Release information should be immutable, any changes should lead to a new release. 

• Processes 

Execute the app as one or more stateless processes. This means that our microservices should be stateless in nature, and should not rely on any state being present in memory or in the filesystem. Indeed the state does not belong in the code. So no sticky sessions, no in-memory cache, no local filesystem storage, etc. Distributed cache like memcache, ehcache or Redis should be used instead

• Port Binding 

Export services via port binding. This is about having your application as standalone, instead of relying on a running instance of an application server, where you deploy. Spring boot provides a mechanism to create a self-executable uber jar that contains all dependencies and embedded servlet container (jetty or tomcat). 

• Concurrency 

Scale-out via the process model. In the twelve-factor app, processes are a first-class citizen. This does not exclude individual processes from handling their own internal multiplexing, via threads inside the runtime VM, or the async/evented model found in tools such as EventMachine, Twisted, or Node.js. But an individual VM can only grow so large (vertical scale), so the application must also be able to span multiple processes running on multiple physical machines. Twelve-factor app processes should never write PID files, rather it should rely on operating system process manager such as systemd - a distributed process manager on a cloud platform. 

• Disposability 

The twelve-factor app’s processes are disposable, meaning they can be started or stopped at a moment’s notice. This facilitates fast elastic scaling, rapid deployment of code or config changes, and robustness of production deploys. Processes should strive to minimize startup time. Ideally, a process takes a few seconds from the time the launch command is executed until the process is up and ready to receive requests or jobs. Short startup time provides more agility for the release process and scaling up; and it aids robustness because the process manager can more easily move processes to new physical machines when warranted. 

• Dev/Prod parity 

Keep development, staging, and production as similar as possible. Your development environment should almost identical to a production one (for example, to avoid some “works on my machine” issues). That doesn’t mean your OS has to be the OS running in production, though. Docker can be used for creating logical separation for your microservices. 

• Logs 

Treat logs as event streams, sending all logs only to stdout. Most Java Developers would not agree to this advice, though. 

• Admin processes 

Run admin/management tasks as one-off processes. For example, a database migration should be run using a separate process altogether.

1. DevOps is a must, because of the explosion of a number of processes in a production system. How to start and stop the fleet of services?
2. The complexity of distributed computing such as “network latency, fault tolerance, message serialization, unreliable networks, handling asynchronous o/p, varying loads within our application tiers, distributed transactions, etc.”
3. How to make configuration changes across the large fleet of services with minimal effort?
4. How to deploy multiple versions of single microservice and route calls appropriately?
5. How to disconnect a microservice from ecosystem when it starts to crash unexpectedly?
6. How to isolate a failed microservice and avoid cascading failures in the entire ecosystem?
7. How to discover services in an elastic manner considering that services may be going UP or DOWN at any point in time?
8. How to aggregate logs/metrics across the services? How to identify different steps of a single client request spread across a span of microservices?

Microservices architecture is meant for developing large distributed systems that scale with safely. There are many benefits of microservices architecture over monoliths, for example:

1. Monolith application is built as a single unit, it is usually composed of 3 components – a database (usually a RDBMS), a server-side executable (war file deployed in tomcat, websphere etc) and a client interface (JSP, etc.)
2. Whenever we want to add/update functionality, developers need to change at least one of these three components and deploy the new version to production. The entire system is tightly coupled, have limitations in choosing technology stack, have low cohesion. 
3. When we need to scale a monolith, we deploy the same version of the monolith on multiple machines, by copying the big war/ear file again and again. Everything is contained into a single executable file.
4. Microservices Architecture, on the other hand, is composed of small autonomous services, divided over business capabilities that communicate with each other over network mostly in async fashion. 

As illustrated in the above example, a typical monolith eShop application is usually a big war file deployed in a single JVM process (tomcat/jboss/websphere, etc). Different components of a monolith communicate with each other using in-process communication like direct method invocation. One or more databases are shared among different components of a monolith application.

Microservices should be autonomous and divided based on business capabilities. Each software component should have single well-defined responsibility (a.k.a Single Responsibility Principle) and the principle of Bounded Context (as defined by Domain Driven Design) should be used to create highly cohesive software components. 

For example, an e-commerce site can be partitioned into following microservices based on its business capabilities:

• Product catalogue
  

Responsible for product information, searching products, filtering products & products facets.

• Inventory

Responsible for managing inventory of products (stock/quantity and facet). 

• Product review and feedback
  

Collecting feedback from users about the products.

• Orders

Responsible for creating and managing orders.

• Payments
  

Process payments both online and offline (Cash On Delivery).

• Shipments

Manage and track shipments against orders.

• Demand generation
  

Market products to relevant users.

• User Accounts
  

Manage users and their preferences.

• Recommendations

Suggest new products based on the user’s preference or past purchases.

• Notifications
  

Email and SMS notification about orders, payments, and shipments.

The client application (browser, mobile app) will interact with these services via API gateway and render the relevant information to the user.

If you want to halt the service when it is not able to locate the config-server during bootstrap, then you need to configure the following property in microservice’s bootstrap.yml: 

spring:
      cloud:
         config:
             fail-fast: true 

Using this configuration will make microservice startup fail with an exception when config-server is not reachable during bootstrap.

We can enable a retry mechanism where microservice will retry 6 times before throwing an exception. We just need to add spring-retry and spring-boot-starter-aop to the classpath to enable this feature. 

build.gradle:- 

...
 dependencies {
   compile('org.springframework.boot:spring-boot-starter-aop')
   compile('org.springframework.retry:spring-retry')
   ...
}

Microservices should be autonomous and divided based on business capabilities. Each software component should have single well-defined responsibility (a.k.a Single Responsibility Principle) and the principle of Bounded Context (as defined by Domain Driven Design) should be used to create highly cohesive software components. 

For example, an e-shop can be partitioned into following microservices based on its business capabilities:

• Product catalogue

Responsible for product information, searching products, filtering products & products facets.

• Inventory 

Responsible for managing inventory of products (stock/quantity and facet).

• Product review and feedback 

Collecting feedback from users about the products.

• Orders 
  

Responsible for creating and managing orders.

• Payments

Process payments both online and offline (Cash On Delivery).

• Shipments 

Manage and track shipments against orders. 

• Demand generation 
  

Market products to relevant users. 

• User Accounts 
  

Manage users and their preferences. 

• Recommendations 
  

Suggest new products based on the user’s preference or past purchases. 

• Notifications 

Email and SMS notification about orders, payments, and shipments.
The client application (browser, mobile app) will interact with these services via the API gateway and render the relevant information to the user.

A good, albeit non-specific, rule of thumb is as small as possible but as big as necessary to represent the domain concept they own said by Martin Fowler

Size should not be a determining factor in microservices, instead bounded context principle and single responsibility principle should be used to isolate a business capability into a single microservice boundary.

Microservices are usually small but not all small services are microservices. If any service is not following the Bounded Context Principle, Single Responsibility Principle, etc. then it is not a microservice irrespective of its size. So the size is not the only eligibility criteria for a service to become microservice.

In fact, size of a microservice is largely dependent on the language (Java, Scala, PHP) you choose, as few languages are more verbose than others.

Microservices are often integrated using a simple protocol like REST over HTTP. Other communication protocols can also be used for integration like AMQP, JMS, Kafka, etc.

The communication protocol can be broadly divided into two categories- synchronous communication and asynchronous communication.

• Synchronous Communication

RestTemplate, WebClient, FeignClient can be used for synchronous communication between two microservices. Ideally, we should minimize the number of synchronous calls between microservices because networks are brittle and they introduce latency. Ribbon - a client-side load balancer can be used for better utilization of resource on the top of RestTemplate. Hystrix circuit breaker can be used to handle partial failures gracefully without a cascading effect on the entire ecosystem. Distributed commits should be avoided at any cost, instead, we shall opt for eventual consistency using asynchronous communication.

• Asynchronous Communication
  

In this type of communication, the client does not wait for a response, instead, it just sends the message to the message broker. AMQP (like RabbitMQ) or Kafka can be used for asynchronous communication across microservices to achieve eventual consistency.

1. You must use asynchronous communication while handling HTTP POST/PUT (anything that modifies the data) requests, using some reliable queue mechanism (RabbitMQ, AMQP, etc.) 
2. It's fine to use synchronous communication for Aggregation pattern at API Gateway Level. But this aggregation should not include any business logic other than aggregation. Data values must not be transformed at Aggregator, otherwise, it defeats the purpose of Bounded Context. In Asynchronous communication, events should be published into a Queue. Events contain data about the domain, it should not tell what to do (action) on this data. 
3. If microservice to microservice communication still requires synchronous communication for GET operation, then seriously reconsider the partitioning of your microservices for bounded context, and create some tasks in backlog/technical debt.

In Orchestration, we rely on a central system to control and call other Microservices in a certain fashion to complete a given task. The central system maintains the state of each step and sequence of the overall workflow. In Choreography, each Microservice works like a State Machine and reacts based on the input from other parts. Each service knows how to react to different events from other systems. There is no central command in this case.

Orchestration is a tightly coupled approach and is an anti-pattern in a microservices architecture. Whereas, Choreography’s loose coupling approach should be adopted where-ever possible.

Example

Let’s say we want to develop a microservice that will send product recommendation email in a fictitious e-shop. In order to send Recommendations, we need to have access to user’s order history which lies in a different microservices. 

In Orchestration approach, this new microservice for recommendations will make synchronous calls to order service and fetch the relevant data, then based on his past purchases we will calculate the recommendations. Doing this for a million users will become cumbersome and will tightly couple the two microservices. 

In Choreography approach, we will use event-based Asynchronous communication where whenever a user makes a purchase, an event will be published by order service. Recommendation service will listen to this event and start building user recommendation. This is a loosely coupled approach and highly scalable. The event, in this case, does not tell about the action, but just the data.

There is no right answer to this question, there could be a release every ten minutes, every hour or once a week. It all depends on the extent of automation you have at a different level of the software development lifecycle - build automation, test automation, deployment automation and monitoring. And of course on the business requirements - how small low-risk changes you care making in a single release.

In an ideal world where boundaries of each microservices are clearly defined (bounded context), and a given service is not affecting other microservices, you can easily achieve multiple deployments a day without major complexity.

Examples of deployment/release frequency

1. Amazon is on record as making changes to production every 11.6 seconds on average in May of 2011.
2. Github is well known for its aggressive engineering practices, deploying code into production on an average 60 times a day.
3. Facebook releases to production twice a day.
4. Many Google services see releases multiple times a week, and almost everything in Google is developed on mainline.
5. Etsy Deploys More Than 50 Times a Day.

Cloud-Native Applications (NCA) is a style of application development that encourages easy adoption of best practices in the area of continuous delivery and distributed software development. These applications are designed specifically for a cloud computing architecture (AWS, Azure, CloudFoundary, etc).

DevOps, continuous delivery, microservices, and containers are the key concepts in developing cloud-native applications.

Spring Boot, Spring Cloud, Docker, Jenkins, Git are a few tools that can help you write Cloud-Native Application without much effort.

• Microservices

It is an architectural approach for developing a distributed system as a collection of small services. Each service is responsible for a specific business capability, runs in its own process and communicates via HTTP REST API or messaging (AMQP).

• DevOps

It is a collaboration between software developers and IT operations with a goal of constantly delivering high-quality software as per customer needs.

• Continuous Delivery

Its all about automated delivery of low-risk small changes to production, constantly. This makes it possible to collect feedback faster.

• Containers

Containers (e.g. Docker) offer logical isolation to each microservices thereby eliminating the problem of "run on my machine" forever. It’s much faster and efficient compared to Virtual Machines.

Spring Boot along with Spring Cloud is a very good option to start building microservices using Java language. There are a lot of modules available in Spring Cloud that can provide boiler plate code for different design patterns of microservices, so Spring Cloud can really speed up the development process. Also, Spring boot provides out of the box support to embed a servlet container (tomcat/jetty/undertow) inside an executable jar (uber jar), so that these jars can be run directly from the command line, eliminating the need of deploying war files into a servlet container. 

You can also use Docker container to ship and deploy the entire executable package onto a cloud environment. Docker can also help eliminate "works on my machine" problem by providing logical separation for the runtime environment during the development phase. That way you can gain portability across on-premises and cloud environment.

Spring Boot makes it easy to create stand-alone, production-grade Spring-based applications that you can "just run" with an opinionated view of the Spring platform and third-party libraries so you can get started with minimum fuss. 

Main features of Spring Boot 

1. Create stand-alone Spring applications (12-factor app style) 
2. Embed Tomcat, Jetty or Undertow directly (no need to deploy WAR files) 
3. Provide opinionated starter POMs to simplify your Maven or Gradle configuration 
4. Automatically configure Spring whenever possible 
5. Provide production-ready features such as metrics, health checks, and externalized configuration 
6. Absolutely no code generation and no requirement for XML configuration 

You can create a Spring Boot starter project by selecting the required dependencies for your project using online tool hosted at https://start.spring.io/ 

Bare minimum dependency for any spring boot application is: 

dependencies {
      compile("org.springframework.boot:spring-boot-starter-web:2.0.4.RELEASE")
} 
The main java class for Spring Boot application will look something like the following: 
import org.springframework.boot.*;
import org.springframework.boot.autoconfigure.*; import org.springframework.stereotype.*;
import org.springframework.web.bind.annotation.*; 
@Controller @EnableAutoConfiguration
public class HelloWorldController {
@RequestMapping("/") @ResponseBody
String home() { 
return "Hello World!"; 
} 
public static void main(String[] args) throws Exception { SpringApplication.run(SampleController.class, args); 
} } 

You can directly run this class, without deploying it to a servlet container. 

Useful References 

• Spring Boot Project
• Spring Boot Starter
• Building an Application with Spring Boot

API Gateway is a special class of microservices that meets the need of a single client application (such as android app, web app, angular JS app, iPhone app, etc) and provide it with single entry point to the backend resources (microservices), providing cross-cutting concerns to them such as security, monitoring/metrics & resiliency. 

Client Application can access tens or hundreds of microservices concurrently with each request, aggregating the response and transforming them to meet the client application’s needs. Api Gateway can use a client-side load balancer library (Ribbon) to distribute load across instances based on round-robin fashion. It can also do protocol translation i.e. HTTP to AMQP if necessary. It can handle security for protected resources as well.

Features of API Gateway

1. Spring Cloud DiscoveryClient integration
2. Request Rate Limiting (available in Spring Boot 2.x)
3. Path Rewriting
4. Hystrix Circuit Breaker integration for resiliency

As the name suggests, zero-downtime deployments do not bring outage in a production environment. It is a clever way of deploying your changes to production, where at any given point in time, at least one service will remain available to customers.

• Blue-green deployment 
  

One way of achieving this is blue/green deployment. In this approach, two versions of a single microservice are deployed at a time. But only one version is taking real requests. Once the newer version is tested to the required satisfaction level, you can switch from older version to newer version.

You can run a smoke-test suite to verify that the functionality is running correctly in the newly deployed version. Based on the results of smoke-test, newer version can be released to become the live version.

• Changes required in client code to handle zero-downtime 

Lets say you have two instances of a service running at the same time, and both are registered in Eureka registry. Further, both instances are deployed using two distinct hostnames: 

/src/main/resources/application.yml 
  spring.application.name: ticketBooks-service
  ---
  spring.profiles: blue
  eureka.instance.hostname: ticketBooks-service -blue.example.com
  ---
  spring.profiles: green

  eureka.instance.hostname: ticketBooks-service -green.example.com

Now the client app that needs to make api calls to books-service may look like below: 

@RestController 
@SpringBootApplication 
@EnableDiscoveryClient
 public class ClientApp { 
@Bean
@LoadBalanced
public RestTemplate restTemplate() { 
return new RestTemplate(); } 
@RequestMapping("/hit-some-api") 
public Object hitSomeApi() { 
return restTemplate().getForObject("https://ticketBooks-service/some-uri", Object.class);  } 

Now, when ticketBooks-service-green.example.com goes down for upgrade, it gracefully shuts down and delete its entry from Eureka registry. But these changes will not be reflected in the ClientApp until it fetches the registry again (which happens every 30 seconds). So for upto 30 seconds, ClientApp’s @LoadBalanced RestTemplate may send the requests to ticketBooks-service-green.example.com even if its down. 

To fix this, we can use Spring Retry support in Ribbon client-side load balancer. To enable Spring Retry, we need to follow the below steps: 

Add spring-retry to build.gradle dependencies 

compile("org.springframework.boot:spring-boot-starter-aop")
compile("org.springframework.retry:spring-retry")

Now enable spring-retry mechanism in ClientApp using @EnableRetry annotation, as shown below: 

@EnableRetry @RestController @SpringBootApplication @EnableDiscoveryClient public class ClientApp { 
... } 

Once this is done, Ribbon will automatically configure itself to use retry logic and any failed request to ticketBooks-service-green.example.com com will be retried to next available instance (in round-robins fashion) by Ribbon. You can customize this behaviour using the below properties: 

/src/main/resources/application.yml 
ribbon:
MaxAutoRetries: 5 
MaxAutoRetriesNextServer: 5 
OkToRetryOnAllOperations: true
OkToRetryOnAllErrors: true

The deployment scenario becomes complex when there are database changes during the upgrade. There can be two different scenarios: 1. database change is backward compatible (e.g. adding a new table column) 2. database change is not compatible with an older version of the application (e.g. renaming an existing table column) 

1. Backward compatible change: This scenario is easy to implement and can be fully automated using Flyway. We can add the script to create a new column and the script will be executed at the time of deployment. Now during blue/green deployment, two versions of the application (say v1 and v2) will be connected to the same database. We need to make sure that the newly added columns allow null values (btw that’s part of the backward compatible change). If everything goes well, then we can switch off the older version v1, else application v2 can be taken off. 
2. Non-compatible database change: This is a tricky scenario, and may require manual intervention in-case of rollback. Let's say we want to rename first_name column to fname in the database. Instead of directly renaming, we can create a new column fname and copy all existing values of first_name into fname column, keeping the first_name column as it is in the database. We can defer non-null checks on fname to post-deployment success. If the deployment goes successful, we need to migrate data written to first_name by v1 to the new column (fname) manually after bringing down the v1. If the deployment fails for v2, then we need to do the otherwise. 

Complexity may be much more in a realistic production app, such discussions are beyond the scope of this book.

ACID is an acronym for four primary attributes namely atomicity, consistency, isolation, and durability ensured by the database transaction manager. 

• Atomicity 

In a transaction involving two or more entities, either all of the records are committed or none are. 

• Consistency 

A database transaction must change affected data only in allowed ways following specific rules including constraints/triggers etc. 

• Isolation 

Any transaction in progress (not yet committed) must remain isolated from any other transaction. 

• Durability 

Committed records are saved by a database such that even in case of a failure or database restart, the data is available in its correct state. 

In a distributed system involving multiple databases, we have two options to achieve ACID compliance: 

1. One way to achieve ACID compliance is to use a two-phase commit (a.k.a 2PC), which ensures that all involved services must commit to transaction completion or all the transactions are rolled back.
2. Use eventual consistency, where multiple databases owned by different microservices become eventually consistent using asynchronous messaging using messaging protocol. Eventual consistency is a specific form of weak consistency. 

2 Phase Commit should ideally be discouraged in microservices architecture due to its fragile and complex nature. We can achieve some level of ACID compliance in distributed systems through eventual consistency and that should be the right approach to do it.

Spring team has an integrated number of battle-tested open-source projects from companies like Pivotal, Netflix into a Spring project known as Spring Cloud. Spring Cloud provides libraries & tools to quickly build some of the common design patterns of a distributed system, including the following:

Spring Cloud makes it really easy to develop, deploy and operate JVM applications for the Cloud.

It is an important design decision. The communication between services might or might not be necessary. It can happen synchronously or asynchronously. It can happen sequentially or it can happen in parallel. So, once we have decided what should be our communication mechanism, we can decide the technology which suits the best.

Here are some of the examples which you can consider.

• Communication can be done by using some queuing service like rabbitmq, activemq and kafka. This is called asynchronous communication.
• Direct API calls can also be made to microservice. With this approach, interservice dependency increases. This is called synchronous communication.
• Webhooks to push data to connected clients/services.

There are mainly two ways to achieve authentication in microservices architecture.

• Centralized sessions

All the microservices can use a central session store and user authentication can be achieved. This approach works but has many drawbacks as well. Also, the centralized session store should be protected and services should connect securely. The application needs to manage the state of the user, so it is called stateful session.

• Token-based authentication/authorization

In this approach, unlike the traditional way, information in the form of token is held by the clients and the token is passed along with each request. A server can check the token and verify the validity of the token like expiry, etc. Once the token is validated, the identity of the user can be obtained from the token. However, encryption is required for security reasons. JWT(JSON web token) is the new open standard for this, which is widely used. Mainly used in stateless applications. Or, you can use OAuth based authentication mechanisms as well.

Logging is a very important aspect of any application. If we have done proper logging in an application, it becomes easy to support other aspects of the application as well. Like in order to debug the issues / in order to understand what business logic might have been executed, it becomes very critical to log important details.

Ideally, you should follow the following practices for logging.

• In a microservice architecture, each request should have a unique value (correlationid) and this value should be passed to each and every microservice so the correlationid can be logged across the services. Thus the requests can be traced.
• Logs generated by all the services should be aggregated in a single location so that while searching becomes easier. Generally, people use ELK stack for the same. So that it becomes easy for support persons to debug the issue.

Docker helps in many ways for microservices architecture.

1. In a microservice architecture, there can be many different services written in different languages. So a developer might have to setup few services along with its dependency and platform requirements. This becomes difficult with the growing number of services in an ecosystem. However, this becomes very easy if these services run inside a Docker container.
2. Running services inside a container also give a similar setup across all the environments, i.e development, staging and production.
3. Docker also helps in scaling along with container orchestration.
4. Docker helps to upgrade the underlying language very easily. We can save many man-hours.
5. Docker helps to onboard the engineers fast.
6. Docker also helps to reduce the dependencies on IT Teams to set up and manage the different kind of environment required.

As container based deployment involves a single image per microservice, it is a bad idea to bundle the configuration along with the image.

This approach is not at all scalable because we might have multiple environments and also we might have to take care of geographically distributed deployments where we might have different configurations as well.

Also, when there are application and cron application as part of the same codebase, it might need to take additional care on production as it might have repercussions how the crons are architected.

To solve this, we can put all our configuration in a centralized config service which can be queried by the application for all its configurations at the runtime. Spring cloud is one of the example services which provides this facility.

It also helps to secure the information, as the configuration might have passwords or access to reports or database access controls. Only trusted parties should be allowed to access these details for security reasons.

In a production environment, you don’t just deal with the application code/application server. You need to deal with API Gateway, Proxy Servers, SSL terminators, Application Servers, Database Servers, Caching Services, and other dependent services.

As in modern microservice architecture where each microservice runs in a separate container, deploying and managing these containers is very challenging and might be error-prone.

Container orchestration solves this problem by managing the life cycle of a container and allows us to automate the container deployments.

It also helps in scaling the application where it can easily bring up a few containers. Whenever there is a high load on the application and once the load goes down. it can scale down as well by bringing down the containers. It is helpful to adjust cost based on requirements.

Also in some cases, it takes care of internal networking between services so that you need not make any extra effort to do so. It also helps us to replicate or deploy the docker images at runtime without worrying about the resources. If you need more resources, you can configure that in orchestration services and it will be available/deployed on production servers within minutes.

An API Gateway is a service which sits in front of the exposed APIs and acts as an entry point for a group of microservices. Gateway also can hold the minimum logic of routing calls to microservices and also an aggregation of the response.

• A gateway can also authenticate requests by verifying the identity of a user by routing each and every request to authentication service before routing it to the microservice with authorization details in the token.
• Gateways are also responsible to load balance the requests.
• API Gateways are responsible to rate limit a certain type of request to save itself from blocking several kinds of attacks etc.
• API Gateways can whitelist or blacklist the source IP Addresses or given domains which can initiate the call.
• API Gateways can also provide plugins to cache certain type of API responses to boost the performance of the application.

One should avoid sharing database between microservices, instead APIs should be exposed to perform the change.

If there is any dependency between microservices then the service holding the data should publish messages for any change in the data for which other services can consume and update the local state.

If consistency is required then microservices should not maintain local state and instead can pull the data whenever required from the source of truth by making an API call.

In the microservices architecture, it is possible that due to service boundaries, a lot of times you need to update one or more entities on the state change of one of the entities. In that case, one needs to publish a message and new event gets created and appended to already executed events. In case of failure, one can replay all events in the same sequence and you will get the desired state as required. You can think of event sourcing as your bank account statement.

You will start your account with initial money. Then all of the credit and debit events happen and the latest state is generated by calculating all of the events one by one. In a case where events are too many, the application can create a periodic snapshot of events so that there isn’t any need to replay all of the events again and again.

Servers come and go in a cloud environment, and new instances of same services can be deployed to cater increasing load of requests. So it becomes absolutely essential to have service registry & discovery that can be queried for finding address (host, port & protocol) of a given server. We may also need to locate servers for the purpose of client-side load balancing (Ribbon) and handling failover gracefully (Hystrix). 

Spring Cloud solves this problem by providing a few ready-made solutions for this challenge. There are mainly two options available for the service discovery - Netflix Eureka Server and Consul. Let's discuss both of these briefly:

Netflix Eureka Server 

Eureka is a REST (Representational State Transfer) based service that is primarily used in the AWS cloud for locating services for the purpose of load balancing and failover of middle-tier servers. The main features of Netflix Eureka are:

1. It provides service-registry.
2. zone aware service lookup is possible.
3. eureka-client (used by microservices) can cache the registry locally for faster lookup. The client also has a built-in load balancer that does basic round-robin load balancing. 

Spring Cloud provides two dependencies - eureka-server and eureka-client. Eureka server dependency is only required in eureka server’s build.gradle 

• build.gradle - Eureka Server 
• compile('org.springframework.cloud:spring-cloud-starter-netflix-eureka-server')

On the other hand, each microservice need to include the eureka-client dependencies to enables eureka discovery. 

build.gradle - Eureka Client (to be included in all microservices) compile('org.springframework.cloud:spring-cloud-starter-netflix-eureka-client').

Eureka server provides a basic dashboard for monitoring various instances and their health in the service registry. The ui is written in freemarker and provided out of the box without any extra configuration. Screenshot for Eureka Server looks like the following. 

It contains a list of all services that are registered with Eureka Server. Each server has information like zone, host, port, and protocol. 

Consul Server 

It is a REST-based tool for dynamic service registry. It can be used for registering a new service, locating a service and health checkup of a service. 

You have the option to choose any one of the above in your spring cloud-based distributed application. In this book, we will focus more on the Netflix Eureka Server option.

If you have 3 different environments (develop/stage/production) in your project setup, then you need to create three different config storage projects. So in total, you will have four projects: 

• config-server 

It is the config-server that can be deployed in each environment. It is the Java Code without configuration storage. 

• config-dev 

It is the git storage for your development configuration. All configuration related to each microservices in the development environment will fetch its config from this storage. This project has no Java code, and t is meant to be used with config-server.

• config-qa 
  

Same as config-dev but its meant to be used only in qa environment.

• Config-prod
  • Same as config-dev, but meant for production environment.
  • So depending upon the environment, we will use config-server with either config-dev, config-qa or config-prod.

There are two main components in Eureka project: eureka-server and eureka-client.

• Eureka Server

The central server (one per zone) that acts as a service registry. All microservices register with this eureka server during app bootstrap.

• Eureka Client

Eureka also comes with a Java-based client component, the eureka-client, which makes interactions with the service much easier. The client also has a built-in load balancer that does basic round-robin load balancing. Each microservice in the distributed ecosystem much include this client to communicate and register with eureka-server.

• Typical use case for Eureka

There is usually one eureka server cluster per region (US, Asia, Europe, Australia) which knows only about instances in its region. Services register with Eureka and then send heartbeats to renew their leases every 30 seconds. If the service can not renew their lease for a few times, it is taken out of server registry in about 90 seconds. The registration information and the renewals are replicated to all the eureka nodes in the cluster. The clients from any zone can look up the registry information (happens every 30 seconds) to locate their services (which could be in any zone) and make remote calls.

Eureka clients are built to handle the failure of one or more Eureka servers. Since Eureka clients have the registry cache information in them, they can operate reasonably well, even when all of the eureka servers go down.

Microservices often need to make remote network calls to another microservices running in a different process. Network calls can fail due to many reasons, including-

1. Brittle nature of the network itself
2. Remote process is hung or
3. Too much traffic on the target microservices than it can handle

This can lead to cascading failures in the calling service due to threads being blocked in the hung remote calls. A circuit breaker is a piece of software that is used to solve this problem. The basic idea is very simple - wrap a potentially failing remote call in a circuit breaker object that will monitor for failures/timeouts. Once the failures reach a certain threshold, the circuit breaker trips, and all further calls to the circuit breaker return with an error, without the protected call being made at all. This mechanism can protect the cascading effects of a single component failure in the system and provide the option to gracefully downgrade the functionality.

A typical use of circuit breaker in microservices architecture looks like the following diagram-

Typical Circuit Breaker Implementation 

Here a REST client calls the Recommendation Service which further communicates with Books Service using a circuit breaker call wrapper. As soon as the books-service API calls starts to fail, circuit breaker will trip (open) the circuit and will not make any further call to book-service until the circuit is closed again. 

Martin Fowler has beautifully explained this phenomenon in detail on his blog. 

Martin Fowler on Circuit Breaker Pattern : https://martinfowler.com/bliki/CircuitBreaker.html 

Circuit Breaker wraps the original remote calls inside it and if any of these calls fails, the failure is counted. When the service dependency is healthy and no issues are detected, the circuit breaker is in Closed state. All invocations are passed through to the remote service. 

If the failure count exceeds a specified threshold within a specified time period, the circuit trips into the Open State. In the Open State, calls always fail immediately without even invoking the actual remote call. The following factors are considered for tripping the circuit to Open State - 

• An Exception thrown (HTTP 500 error, can not connect)
• Call takes longer than the configured timeout (default 1 second)
• The internal thread pool (or semaphore depending on configuration) used by hystrix for the command execution rejects the execution due to exhausted resource pool. 

After a predetermined period of time (by default 5 seconds), the circuit transitions into a half-open state. In this state, calls are again attempted to the remote dependency. Thereafter the successful calls transition the circuit breaker back into the closed state, while the failed calls return the circuit breaker into the open state.

1. Synchronous communication over the network that is likely to fail is a potential candidate for circuit breaker.
2. A circuit breaker is a valuable place for monitoring, any change in the breaker state should be logged so as to enable deep monitoring of microservices. It can easily troubleshoot the root cause of failure.
3. All places where a degraded functionality can be acceptable to the caller if the actual server is struggling/down. 

Benefits:-

1. The circuit breaker can prevent a single service from failing the entire system by tripping off the circuit to the faulty microservice. 
2. The circuit breaker can help to offload requests from a struggling server by tripping the circuit, thereby giving it a time to recover. 
3. In providing a fallback mechanism where a stale data can be provided if real service is down.

Config first bootstrap and discovery first bootstrap are two different approaches for using Spring Cloud Config client in Spring Cloud-powered microservices. Let’s discuss both of them:

Config First Bootstrap 

This is the default behavior for any spring boot application where Spring Cloud Config client is on the classpath. When a config client starts up it binds to the Config Server using the bootstrap configuration property and initializes Spring Environment with remote property sources.

• Config-first approach 

The only configuration that each microservice (except config-server) needs to provide is the following: 

• File:-  /src/main/resources/bootstrap.yml 
• spring.cloud.config.uri: http://localhost:8888 
• In config-first approach, even the eureka-server can fetch its own configuration from config-server. Point worth noting down here is that config-server must be the first service to boot up in the entire ecosystem, because each service will fetch its configuration from config-server. 

Discovery First Bootstrap 

If you are using Spring Cloud Netflix and Eureka Service Discovery then you can have Config Server register with the Discovery Service and let all clients get access to config server via discovery service. 

• Discovery-first approach 

This is not the default configuration in Spring Cloud applications, so we need to manually enable it using the below property in bootstrap.yml 

Listing 17. /src/main/resources/bootstrap.yml 
  spring:
      cloud:
          config:
              discovery:
enabled: true

This property should be provided by all microservices so that they can take advantage of discovery first approach. 

The benefit of this approach is that now config-server can change its host/port without other microservices knowing about it since each microservice can get the configuration via eureka service now. The downside of this approach is that an extra network round trip is required to locate the service registration at app startup. 

Strangulation is used to slowly decommission an older system and migrate the functionality to a newer version of microservices. 

Normally one endpoint is Strangled at a time, slowly replacing all of them with the newer implementation. Zuul Proxy (API Gateway) is a useful tool for this because we can use it to handle all traffic from clients of the old endpoints, but redirect only selected requests to the new ones. 

Let’s take an example use-case: 

/src/main/resources/application.yml 
zuul:
    routes:

first:

path: /first/**
url: http://first.example.com --1 

legacy:

path: /**
url: http://legacy.example.com  -- 2 

1. Paths in /first/** have been extracted into a new service with an external URL http://first.example.com 
2. legacy app is mapped to handle all request that do not match any other patterns (/first/**). 

This configuration is for API Gateway (zuul reverse proxy), and we are strangling selected endpoints /first/ from the legacy app hosted at http://legacy.example.com slowly to newly created microservice with external URL http://first.example.com

Hystrix is Netflix implementation for circuit breaker pattern, that also employs bulkhead design pattern by operating each circuit breaker within its own thread pool. It also collects many useful metrics about the circuit breaker’s internal state, including -

1. Traffic volume.
2. Request volume.
3. Error percentage.
4. Hosts reporting
5. Latency percentiles.
6. Successes, failures, and rejections.

All these metrics can be aggregated using another Netflix OSS project called Turbine. Hystrix dashboard can be used to visualize these aggregated metrics, providing excellent visibility into the overall health of the distributed system.
Hystrix can be used to specify the fallback method for execution in case the actual method call fails. This can be useful for graceful degradation of functionality in case of failure in remote invocation. 

Add hystrix library to build.gradle dependencies { 
compile('org.springframework.cloud:spring-cloud-starter-hystrix') 

1) Enable Circuit Breaker in main application 

@EnableCircuitBreaker @RestController @SpringBootApplication
public class ReadingApplication {
... } 

2) Using HystrixCommand fallback method execution 

@HystrixCommand(fallbackMethod = "reliable") 
public String readingList() {
URI uri = URI.create("http://localhost:8090/recommended"); return this.restTemplate.getForObject(uri, String.class); 
} 
public String reliable() { 2
return "Cached recommended response"; 
}

1. Using @HystrixCommand annotation, we specify the fallback method to execute in case of exception. 
2. fallback method should have the same signature (return type) as that of the original method. This method provides a graceful fallback behavior while the circuit is in the open or half-open state.

Hystrix library makes our distributed system resilient (adaptable & quick to recover) to failures. It 

provides three main features: 

• Latency and fault-tolerance 

It helps stop cascading failures, provide decent fallbacks and graceful degradation of service functionality to confine failures. It works on the idea of fail-fast and rapid recovery. Two different options namely Thread isolation and Semaphore isolation are available for use to confine failures. 

• Real-time operations 

Using real-time metrics, you can remain alert, make decisions, affect changes and see results. 

• Concurrency 

Parallel execution, concurrent aware request caching and finally automated batching through request collapsing improves the concurrency performance of your application. 

More information on Netflix hystrix library: 

• https://github.com/Netflix/Hystrix/
• https://github.com/Netflix/Hystrix/wiki#principles
• https://github.com/Netflix/Hystrix/wiki/How-it-Works

Let's say we want to handle service to service failure gracefully without using the Circuit Breaker pattern. The naive approach would be to wrap the   REST call in a try-catch clause. But Circuit Breaker does a lot more than try-catch can not accomplish - 

1. Circuit Breaker does not even try calls once the failure threshold is reached, doing so reduces the number of network calls. Also, a number of threads consumed in making faulty calls are freed up.
2. Circuit breaker provides fallback method execution for gracefully degrading the behavior. Try catch approach will not do this out of the box without additional boiler plate code.
3. Circuit Breaker can be configured to use a limited number of threads for a particular host/API, doing so brings all the benefits of bulkhead design pattern. 

So instead of wrapping service to service calls with try/catch clause, we must use the circuit breaker pattern to make our system resilient to failures.

The bulkhead implementation in Hystrix limits the number of concurrent calls to a component/service. This way, the number of resources (typically threads) that are waiting for a reply from the component/service is limited.

Let's assume we have a fictitious web e-commerce application as shown in the figure below. The WebFront communicates with 3 different components using remote network calls (REST over HTTP). 

• Product catalogue Service
• Product Reviews Service
• Order Service

Now let's say due to some problem in Product Review Service, all requests to this service start to hang (or timeout), eventually causing all request handling threads in WebFront Application to hang on waiting for an answer from Reviews Service. This would make the entire WebFront Application non-responsive. The resulting behavior of the WebFront Application would be same if request volume is high and Reviews Service is taking time to respond to each request.

The Hystrix Solution

Hystrix’s implementation for bulkhead pattern would limit the number of concurrent calls to components and would have saved the application in this case by gracefully degrading the functionality. Assume we have 30 total request handling threads and there is a limit of 10 concurrent calls to Reviews Service. Then at most 10 request handling threads can hang when calling Reviews Service, the other 20 threads can still handle requests and use components Products and Orders Service. This will approach will keep our WebFront responsive even if there is a failure in Reviews Service. 

Martin Fowler introduced the concept of "smart endpoints & dumb pipes" while describing microservices architecture.

To give context, one of the main characteristic of a   based system is to build small utilities and connect them using pipes. For example, a very popular way of finding all java processes in Linux system is Command pipeline in Unix shell ps elf | grep java

Here two commands are separated by a pipe, the pipe’s job is to forward the output of the first command as an input to the second command, nothing more.   like a dumb pipe which has no business logic except the routing of data from one utility to another.

In his article Martin Fowler compares Enterprise Service Bus (ESB) to ZeroMQ/RabbitMQ, ESB is a pipe but has a lot of logic inside it while ZeroMQ has no logic except the persistence/routing of messages. ESB is a fat layer that does a lot of things like - security checks, routing, business flow & validations, data transformations, etc. So ESB is a kind of smart pipe that does a lot of things before passing data to next endpoint (service). Smart endpoints & dumb pipes advocate an exactly opposite idea where the communication channel should be stripped of any business-specific logic and should only distribute messages between components. The components (endpoints/services) should do all the data validations, business processing, security checks, etc on those incoming messages. 

Microservices team should follow the principles and protocols that worldwide web & Unix is built on.

There are different ways to handle the versioning of your REST api to allow older consumers to still consume the older endpoints. The ideal practice is that any nonbackward compatible change in a given REST endpoint shall lead to a new versioned endpoint. 

Different mechanisms of versioning are: 

• Add version in the URL itself
• Add version in API request header 

Most common approach in versioning is the URL versioning itself. A versioned URL looks like the following: 

Versioned URL 

• https://<host>:<port>/api/v1/...
• https://<host>:<port>/api/v2/...

As an API developer you must ensure that only backward-compatible changes are accommodated in a single version of URL. Consumer-Driven-Tests can help identify potential issues with API upgrades at an early stage.

Using config-server, it's possible to refresh the configuration on the fly. The configuration changes will only be picked by Beans that are declared with @RefreshScope annotation.

The following code illustrates the same. The property message is defined in the config-server and changes to this property can be made at runtime without restarting the microservices. 

package hello;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication; 
import org.springframework.cloud.context.config.annotation.RefreshScope; 
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RestController;
@SpringBootApplication
public class ConfigClientApplication {
public static void main(String[] args) { SpringApplication.run(ConfigClientApplication.class, args);
} } 
@RefreshScope 1 @RestController
class MessageRestController {
@Value("${message:Hello World}") private String message;
@RequestMapping("/message") String getMessage() {
return this.message; }}

1 @RefreshScope makes it possible to dynamically reload the configuration for this bean.

@HystrixCommand annotation provides attribute ignoreExceptions that can be used to provide a list of ignored exceptions.

Code

• import com.netflix.hystrix.contrib.javanica.annotation.HystrixCommand;
• import org.springframework.beans.factory.annotation.Autowired;
• import org.springframework.cloud.client.ServiceInstance;
• import org.springframework.cloud.client.loadbalancer.LoadBalancerClient; 
• import org.springframework.stereotype.Service;
• import org.springframework.web.client.RestTemplate; 
• import java.net.URI;

@Service

public class HystrixService { 

@Autowired

private LoadBalancerClient loadBalancer;

@Autowired

private RestTemplate restTemplate;

@HystrixCommand(fallbackMethod = "reliable", ignoreExceptions = IllegalStateException.class, MissingServletRequestParameterException.class, TypeMismatchException.class)

public String readingList() {

ServiceInstance instance = loadBalancer.choose("product-service"); URI uri = URI.create("http://product-service/product/recommended"); return this.restTemplate.getForObject(uri, String.class);}

public String reliable(Throwable e) { return "Cloud Native Java (O'Reilly)"; 

In the above example, if the actual method call throws IllegalStateException, MissingServletRequestParameterException or TypeMismatchException then hystrix will not trigger the fallback logic (reliable method), instead the actual exception will be wrapped inside HystrixBadRequestException and re-thrown to the caller. It is taken care by javanica library under the hood.

In a microservices architecture, each microservice shall own its private data which can only be accessed by the outside world through owning service. If we start sharing microservice’s private datastore with other services, then we will violate the principle of Bounded Context. 

Practically we have three approaches -

1. Database server per microservice - Each microservice will have its own database server instance. This approach has the overhead of maintaining database instance and its replication/backup, hence its rarely used in a practical environment. 
2. Schema per microservice - Each microservice owns a private database schema which is not accessible to other services. Its most preferred approach for RDMS database (MySql, Postgres, etc.)
3. Private Table per microservice - Each microservice owns a set of tables that must only be accessed by that service. It’s a logical separation of data. This approach is mostly used for the hosted database as a service solution (Amazon RDS).

Microservices Architecture can become cumbersome & unmanageable if not done properly. There are best practices that help design a resilient & highly scalable system. The most important ones are 

• Partition correctly 

Get to know the domain of your business, that's very very important. Only then you will be able to define the bounded context and partition your microservice correctly based on business capabilities. 

• DevOps culture 

Typically, everything from continuous integration all the way to continuous delivery and deployment should be automated. Otherwise,   a big pain to manage a large fleet of microservices. 

• Design for stateless operations 

We never know where a new instance of a particular microservice will be spun up for scaling out or for handling failure, so maintaining a state inside service instance is a very bad idea. 

• Design for failures 

Failures are inevitable in distributed systems, so we must design our system for handling failures gracefully. failures can be of different types and must be dealt with accordingly, for example - 

1. Failure could be transient due to inherent brittle nature of the network, and the next retry may succeed. Such failures must be protected using retry operations.
2. Failure may be due to a hung service which can have cascading effects on the calling service. Such failures must be protected using Circuit Breaker Patterns. A fallback mechanism can be used to provide degraded functionality in this case.
3. A single component may fail and affect the health of the entire system, bulkhead pattern must be used to prevent the entire system from failing. 

• Design for versioning 

We should try to make our services backward compatible, explicit versioning must be used to cater different versions of the RESt endpoints. 

• Design for asynchronous communication b/w services 

Asynchronous communication should be preferred over synchronous communication in inter microservice communication. One of the biggest advantages of using asynchronous messaging is that the service does not block while waiting for a response from another service. 

• Design for eventual consistency 

Eventual consistency is a consistency model used in distributed computing to achieve high availability that informally guarantees that, if no new updates are made to a given data item, eventually all accesses to that item will return the last updated value. 

• Design for idempotent operations 

Since networks are brittle, we should always design our services to accept repeated calls without any side effects. We can add some unique identifier to each request so that service can ignore the duplicate request sent over the network due to network failure/retry logic. 

• Share as little as possible 
  

In monolithic applications, sharing is considered to be a best practice but that's not the case with Microservices. Sharing results in a violation of Bounded Context Principle, so we shall refrain from creating any single unified shared model that works across microservices. For example, if different services need a common Customer model, then we should create one for each microservice with just the required fields for a given bounded context rather than creating a big model class that is shared in all services. 

The more dependencies we have between services, the harder it is to isolate the service changes, making it difficult to make a change in a single service without affecting other services. Also, creating a unified model that works in all services brings complexity and ambiguity to the model itself, making it hard for anyone to understand the model.

In a way are want to violate the DRY principle in microservices architecture when it comes to domain models.

Caching is a technique of performance improvement for getting query results from a service. It helps minimize the calls to network, database, etc. We can use caching at multiple levels in microservices architecture - 

1. Server-Side Caching - Distributed caching software like Redis/MemCache/etc are used to cache the results of business operations. The cache is distributed so all instances of a microservice can see the values from the shared cache. This type of caching is opaque to clients.
2. Gateway Cache - central API gateway can cache the query results as per business needs and provide improved performance. This way we can achieve caching for multiple services at one place. Distributed caching software like Redis or Memcache can be used in this case.
3. Client-Side Caching - We can set cache-headers in http response and allow clients to cache the results for a pre-defined time. This will drastically reduce the load on servers since the client will not make repeated calls to the same resource. Servers can inform the clients when information is changed, thereby any changes in the query result can also be handled. E-Tags can be used for client-side load balancing. If the end client is a microservice itself, then Spring Cache support can be used to cache the results locally. 

Swagger is a very good open-source tool for documenting   APIs provided by microservices. It provides very easy to use interactive documentation.

By the use of swagger annotation on REST endpoint, api documentation can be auto-generated and exposed over the web interface. An internal and external team can use web interface, to see the list of APIs and their inputs & error codes. They can even invoke the endpoints directly from web interface to get the results.

Swagger UI is a very powerful tool for your microservices consumers to help them understand the set of endpoints provided by a given microservice.

Basic Authentication is natively supported by almost all servers and clients, even Spring security has very good support for it and its configured out of the box. But it is not a good fit for Microservices due to many reasons, including -

1. We need credentials (username and password) every time we authenticate. This may be fine where all the participants can share the secrets securely, but Users may not be willing to share their credentials with all the applications.
2. There is no distinction between Users and Client Apps (an application that is making a request). In a realistic environment, we often need to know if a real user is making a request or a client app is making a request (for inter-service communication).
3. It only covers authentication. what about scopes, Authorizations? Basic Auth does not support adding additional attributes in the authentication headers. There is no concept of Tokens in basic auth.
4. Performance reasons for BCrypt Matching. Passwords are often stored in the database using one-way hash i.e. Bcrypt, it takes a lot of cpu cycles depending upon the strength (a.k.a. log rounds in BCrypt) to compare the user’s plain password with db saved bcrypt password, so it may not be efficient to match password on every request. The larger the strength parameter the more work will have to be done (exponentially) to hash the passwords. If you set the strength to 12, then in total 212 iterations will be done in Bcrypt Logic. Usually, 4-8 passwords can be matched per second on a T2.Micro instance on Amazon AWS instance. See BCryptPasswordEncoder for more info.
5. If we use Basic Auth for a mobile application client, then we might have to store user’s credentials on the device to allow remember me feature. This is quite risky as anyone getting access to the device may steal the plain credentials.

There are 3 parts in every JWT claim - Header, Claim and Signature. These 3 parts are separated by a dot. The entire JWT is encoded in Base64 format. 

JWT = {header}.{payload}.{signature} 

A typical JWT is shown here for reference. 

Encoded JSON Web Token

Entire JWT is encoded in Base64 format to make it compatible with HTTP protocol. Encoded JWT looks like the following:

Decoded JSON Web Token 

Header 

Header contains algorithm information e.g. HS256 and type e.g. JWT 

{
"alg": "HS256", "typ": "JWT" 
} 

Claim 

claim part has an expiry, issuer, user_id, scope, roles, client_id etc. It is encoded as a JSON object. You can add custom attributes to the claim. This is the information that you want to exchange with the third party. 

{
"uid": "2ce35360-ef8e-4f69-a8d7-b5d1aec78759", "user_name": "user@mail.com",
"scope": ["read"],
"exp": 1520017228,
"authorities": ["ROLE_USER","ROLE_ADMIN"], "jti": "5b42ca29-8b61-4a3a-8502-53c21e85a117", "client_id": "acme-app" 
} 

Signature 

Signature is typically a one way hash of (header + payload), is calculated using HMAC SHA256 algorithm. The secret used for signing the claim should be kept private. Pubic/private key can also be used to encrypt the claim instead of using symmetric cryptography. 

HMACSHA256(base64(header) + "." + base64(payload), "secret")

OAuth2.0 is a delegation protocol where the Client (Mobile App or web app) does not need to know about the credentials of Resource Owner (end-user). 

Oauth2 defines four roles. 

1. Resource Owner - The person or the application that owns the data to be shared. When a resource owner is a person, it is called as an end-user.
2. Resource Server - The application that holds the protected resources. It is usually a microservice.
3. Authorization Server - the application that verifies the identity of the resource owner (users/clients). These server issues access tokens after obtaining the authorization.
4. Client - the application that makes a request to Resource Server on behalf of Resource Owner. It could be a mobile app or a web app (like stackoverflow).

Important Tools and Libraries for testing Spring-based Microservices are - 

• JUnit 

the standard test runners 

• TestNG 

the next generation test runner 

• Hemcrest 

declarative matchers and assertions 

• Rest-assured 

for writing REST Api driven end to end tests 

• Mockito 

for mocking dependencies 

• Wiremock 

for stubbing thirdparty services 

• Hoverfly 

Create API simulation for end-to-end tests. 

• Spring Test and Spring Boot Test 

for writing Spring Integration Tests - includes MockMVC, TestRestTemplate, Webclient like features. 

• JSONassert 

An assertion library for JSON. 

• Pact 

The Pact family of frameworks provide support for Consumer Driven Contracts testing. 

• Selenium 

Selenium automates browsers. Its used for end-to-end automated ui testing. 

• Gradle 

Gradle helps build, automate and deliver software, fastr. 

• IntelliJ IDEA 

IDE for Java Development 

• Using spring-boot-starter-test 
  • We can just add the below dependency in project’s build.gradle 
  • testCompile('org.springframework.boot:spring-boot-starter-test')

This starter will import two spring boot test modules spring-boot-test & spring-boot-test- autoconfigure as well as Junit, AssertJ, Hamcrest, Mockito, JSONassert, Spring Test, Spring Boot Test and a number of other useful libraries.

There are many useful scenarios for leveraging the power of JWT-

• Authentication

Authentication is one of the most common scenarios for using JWT, specifically in microservices architecture (but not limited to it). In microservices, the oauth2 server generates a JWT at the time of login and all subsequent requests can include the JWT AccessToken as the means for authentication. Implementing Single Sign-On by sharing JWT b/w different applications hosted in different domains.

• Information Exchange

JWT can be signed, using public/private key pairs, you can be sure that the senders are who they say they are. Hence JWT is a good way of sharing information between two parties. An example use case could be -

1. Generating Single Click Action Emails e.g. Activate your account, delete this comment, add this item to favorites, Reset your password, etc. All required information for the action can be put into JWT.
2. Timed sharing of a file download using a JWT link. Timestamp can be part of the claim, so when the server time is past the time-coded in JWT, the link will automatically expire.