After watching the BUILD conference videos for Azure Service Fabric, I'm left imagining how this might be a good fit for our current microservice-based architecture. There is one thing I'm not entirely sure how I would go about solving, however - the API gateway/proxy.
Consider a less-than-trivial microservice architecture where you have N number of services running within the Azure Service Fabric exposing REST endpoints. In many situations, you want to package these fragmented API endpoints up into a single-entry API for consumers to use, to avoid having them connecting to the service fabric-instances directly. The Azure Service Fabric solution seems so complete in every way that I'm sort of wondering if I missed something obvious when I don't see a way to trivially solve this within the capabilities mentioned during the BUILD talks.
Services like Vulcan aim to solve this problem by having the services register the paths they want routed to them in etcd. I'm guessing one way of solving this may be to create a separate stateful web service that other services can register themselves with, providing service name and the paths they need routed to them. The stateful web service can then route traffic to the correct instance based on its state. This doesn't seem entirely ideal, though, with stuff like removing routes when applications are removed and generally keeping the state in sync with the services deployed within the cluster. Has anybody given this any thought, or have any ideas how one might go about solving this within Azure Service Fabric?
The service registration/discoverability you need to do this is actually already there. There's a stateful system service called the Naming Service, which is basically a registrar of service instances and the endpoints they're listening on. So when you start up a service - either stateless or stateful - and open some listener on it, the address gets registered with the Naming Service.
Now the part you'd need to fill in is the "gateway" that users interact with. This doesn't have to be stateful because the Naming Service manages the stateful part. But you'd have to come up with an addressing scheme that works for you, and then it would just forward requests along to the right place. Basically something like this:
Receive request.
Use NS to find the service that can take the request.
Forward the request to it and the response back to the user.
If the service doesn't exist anymore, 404.
In general we don't like to dictate anything about how your services talk to each other, but we are thinking of ways to solve this problem for HTTP as a complete built-in solution.
We implemented a HTTP gateway service for this purpose as well. To make sure we can have one HTTP gateway for any internal protocol, we implemented the gateway for HTTP based internal services (like ASP.NET WebAPIs) using an ASP.NET 5 middleware. It routes requests from e.g /service to an internal Service Fabric address like fabric:/myapp/myservice by using the ServicePartitionClient and some retry logic from CommunicationClientFactoryBase.
We open-sourced this middleware and you can find it here:
https://github.com/c3-ls/ServiceFabric-HttpServiceGateway
There's also some more documentation in the wiki of the project.
This feature is build in for http endpoints, starting with release 5.0 of service fabric. The documentation is available at https://azure.microsoft.com/en-us/documentation/articles/service-fabric-reverseproxy/
We have used an open source project called Traefik with amazing success. There is an Azure Service Fabric wrapper around it - it's essentially a GoLang exe that is deployed onto the cluster as Managed Executable.
It supports circuit breakers, weighted round robin LB, path & header version routing (this is awesome for hosting multiple API versions), the list goes on. And its got a handy portal to view the config and health stats.
The real power in it lies in how you configure it. It's done via the service itself in the ServiceManifest.xml. This allows you to deploy new services and have them immediately able to be routed to - no need to update a routing table etc.
Example
<StatelessServiceType ServiceTypeName="WebServiceType">
<Extensions>
<Extension Name="Traefik">
<Labels xmlns="http://schemas.microsoft.com/2015/03/fabact-no-schema">
<Label Key="traefik.frontend.rule.example">PathPrefixStrip: /a/path/to/service</Label>
<Label Key="traefik.enable">true</Label>
<Label Key="traefik.frontend.passHostHeader">true</Label>
</Labels>
</Extension>
</Extensions>
</StatelessServiceType>
Highly recommended!
Azure Service Fabric makes it easy to implement the standard architecture for this scenario: a gateway service as a frontend for the clients to connect to and all the N backend services communicating with the front end gateway. There are a few communication API stacks available as part of Service Fabric that make it easy to communicate from clients to services and within services themselves. The communication API stacks provided by Service Fabric hide the details of discovering, connecting and retrying connections so that you can focus on the actual exchange of information. When using the Service Fabric communication APIs the services do not have to implement the mechanism of registering their names and endpoints to a specific routing service except what are the usual steps as part of creating the service itself. The communication APIs take in the service URI and partition key and automatically resolve and connect to the right service instance. This article provides a good starting point to help make a decision with regards to which communication APIs will be best suited for your particular case depending on whether you are using Reliable Actors or Reliable Services, or protocols such as HTTP or WCF, or the choice of programming language that the services are written in. At the end of the article you will find links to more detailed articles and tutorials for different communication APIs. For a tutorial on communication in Web API services see this.
We are using SF with a gateway pattern and about 13 services behind the gateway. We use the built in DNS service that SF provides, see: https://learn.microsoft.com/en-us/azure/service-fabric/service-fabric-dnsservice, this allows the internal service to service calls with known (internal to SF) DNS names, including gateway service to internal services. There are some well known asp.net core gateways (Ocelot, ProxyKit) to use, but we rolled our own. We have an external load balancer to route to multiple gateway instances in SF.
When a service is started, it registers it's endpoint with the fabric naming service. Using the Fabric client APIs you can then ask fabric for the registered endpoints, associated with the registered service name.
So yes, just as you described your case, you would have a gateway that would accept an incoming URI for connection, and then use that path information as the service name lookup, to then create a proxy connection between the incoming request and the actual internal endpoint location.
Looks like the team as posted one the samples that shows how to do this: https://github.com/Azure/servicefabric-samples/tree/master/samples/Services/VS2015/WordCount
Related
At present we have a lot of ASP.net WebAPI service applications hosted on premises. We are planning to move these to Azure AKS. We've identified a lot of common code across these applications which is mostly implemented as ASP.Net reusable middleware components so that the logic is not duplicated in code.
In a K8s environment it makes sense to offload this common functionality to one or more proxy applications which intercepts the requests being forwarded from the ingress to the services (assuming this is the correct approach). Some of the request inspection / manipulation logic is based on the service host and path to be defined in the ingress and even on the headers in the incoming requests.
For e.g. I considered using OAuth2_proxy but found that even though authentication is quite easy to implement, Azure AD group based authorization is impossible to do out of the box with that. So what's the idiomatic way one goes about setting up such a custom proxy application? (I'm familiar with using libraries such as ProxyKit middleware in ASP.Net to develop http proxies.)
One approach that comes to mind is to deploy such proxies as sidecar containers in each service application pod but that would mean there'd be unnecessary resource usage by all such duplicate container instances in each pod. I don't see the benefit over the use of middleware components as mentioned previously. :(
The ideal setup would be ingress --> custom proxy 1 --> custom proxy 2 --> custom proxy n --> service where custom proxies would be separately deployable and scalable.
So after a lot of reading and googling I found that the solution was to use API Gateways that are available as libraries (preferrably based on .Net):
Ocelot placed behind the nginx ingress fits the bill perfectly
Ocelot is a .NET API Gateway. This project is aimed at people using .NET running a micro services / service oriented architecture that need a unified point of entry into their system. However it will work with anything that speaks HTTP and run on any platform that ASP.NET Core supports.
Ocelot is currently used by Microsoft and Tencent.
The custom middleware and header/query/claims transformation solves my problem. Here are some worthy links
Microsoft Docs: Implement API Gateways with Ocelot
Ocelot on Github
Ocelot Documentation
Features
A quick list of Ocelot's capabilities for more information see the documentation.
Routing
Request Aggregation
Service Discovery with Consul & Eureka
Service Fabric
Kubernetes
WebSockets
Authentication
Authorisation
Rate Limiting
Caching
Retry policies / QoS
Load Balancing
Logging / Tracing / Correlation
Headers / Query String / Claims Transformation
Custom Middleware / Delegating Handlers
Configuration / Administration REST API
Platform / Cloud Agnostic
I am developing a website using which is based on a microservices architecture which is containerized with docker. Now I want to communicate with microservices using rest API. I read some articles which show some diagrams that API gateway make some connections to microservices. I am confused about the main purpose of the API gateway.
Gateways are used for cross cutting concerns like authentication, logging and redirect to respective backend services. They are usual the single point of entry for applications and are exposed publicly. the rest of the services can sit behind the firewall that your gateway has access to making backend services secure. you also write your resiliency logic in the gateway by implementing retry or circuit breaker patterns etc.
Since gateway has advantages but it can be single point of failure. so make sure it's highly available by deploying redundant copies
I have setup 2 noted type cluster on FrontEnd and another BackEnd. The FrontEnd has stateless services and the Backend has Statefull and actor services. Now I have seen examples where they use reverse proxy and http:// calls to communicate with stateful services, and other places where they use Remote calls calling fabric:// When should each be used if there is data intensive transfers happening between Frontend and BackEnd node types which would be better protocol?
Actually fabric:// isn't a protocol itself, it's just a syntax for Service Fabric Naming Service to resolve actual location of your service. Remoting is a better choice if you don't have to expose your service to external clients, since it will choose protocol depending on locations on client and service (may use interprocess communication in case both are located on the same node) while using http:// sticks you to only this protocol.
The fabric:// is just an Uri Scheme. It is used to identify named services, like: fabric://MyApp/MyService
There is no right answer to this question, there are many variables to take into account to select the right approach.
You can use both and it will be absolutely fine.
It far more than that, but a a simple overview I can give is:
Using HTTP communication, the services depends only on each other endpoints, and both can be treated isolated from each other during development and deployment, they will communicate even when you change services versions and tech-stack. You can use different technologies like: Java, GO, NodeJS and still have a smooth communication between your services.
Using Remoting, you might get faster communication, but higher coupling between the services, because both need to understand same interfaces and entities used for communication, to keep them in sync(compatible) will most of the time require to deploy new version of both services together.
.
If performance is not an issue at start, I would suggest go simple with HTTP, and migrate if does not attend your demands.
I created a secure service fabric and have full control. Only services developed by me will be deployed and all public facing apis are using https. The question is for communication between service to service in cluster, is it safe to use http instead of https?
Thanks,
This is related to how do you set up your cluster with your network. Public facing services and internal services should be in a different NSG networks. You should control very strict communication between public and internal. If you can do that then http or any protocol between services is not a problem
Personally, i prefer http over remoting because there will be less dependency between services. Communication between services can be defined using Open API and therefore become vendor neutral.
There is a very good course on MVA for how to set up Service fabric class and best practice
https://mva.microsoft.com/en-US/training-courses/16925?l=E5B3nJSGD_805167344
I recommend using service fabric remoting (instead of http). This can be secured using certificates, with little effort.
More info here: https://learn.microsoft.com/en-us/azure/service-fabric/service-fabric-reliable-services-secure-communication
Here is usecase:
I have two apps in Bluemix: app1 and app2
app1 is accessible through the internet using its route (e.g. app1.mybluemix.net)
app2 doesn't have any route to prevent from being accessible through the internet.
app2 may expose a REST API.
How do I call app2 from app1 inside Bluemix?
An example of communicating to an application without a route is implemented in this Microservice Shipping sample.
This is an EJB Liberty application that runs on Bluemix without a route and subscribes to the Bluemix MQ Light service. The sender of the messages is the Microservice Orders sample application, which binds to the same MQ Light service.
Going the REST API route will mean you must have an externally accessible route. However, you could secure it using keys and tokens.
It would be easier to use one of the services in Bluemix as an "RPC" layer between the two applications. You could use one of the queue services (MQLight, RabbitMQ) or Redis to pass messages between the applications to execute commands.
These service bindings are internal and won't be exposed externally unlike the REST API.
Alternatively, you could expose the REST API from App2 and use authentication to control access.
There are two ways you can prevent access.
Put your microservice inside a Bluemix Container and utilize private IPs https://new-console.ng.bluemix.net/docs/containers/container_security_network.html#container_cli_ips_byoip
Use API Connect as a API Gateway/proxy to the private IP being in your container microservice.
Use Bluemix Dedicated to deploy app2. Bluemix dedicated provides firewall capabilities and you could set it up so that it only accepts requests from app1's IP address.
Use Bluemix Local when it becomes available with the same approach where you use your corporate firewall to only accept requests that come from your App1 IP Address. This is an expensive alternative compared to a public PAAS.
Use the API Connect Service which replaced the API Management Service to:
Specify what users can access your apis
Specify the number of requests per day or other unit of time
Provides a API Gateway to securely call the other service App2.
I expect at some point a software network defined solution will be considered as part of the offering.