If you plan on developing applications using Steeltoe, you will need to download and install the latest .NET Core SDK.
Note: Many of the Steeltoe packages can also be used with .NET Framework and ASP.NET 4 based applications. You are not required to target .NET/ASP.NET Core when using Steeltoe.
When developing applications using Steeltoe, whether on .NET Core or .NET Framework, you will need to pull the Steeltoe NuGet packages into your application.
To use the latest releases of Steeltoe, you can subscribe to any one of the following feeds, depending on your needs:
Below is an example NuGet.config file you can edit and use when developing applications using Steeltoe. Note that if you only want to use release or pre-release (RC) versions of Steeltoe, you do not need to make any changes to your NuGet.config file, as those are served from
<?xml version="1.0" encoding="utf-8"?> <configuration> <packageSources> <add key="SteeltoeMaster" value="https://www.myget.org/F/steeltoemaster/api/v3/index.json" /> <add key="SteeltoeDev" value="https://www.myget.org/F/steeltoedev/api/v3/index.json" /> <add key="NuGet" value="https://api.nuget.org/v3/index.json" /> </packageSources> </configuration>
For many of the Steeltoe services, we provide Quick Start samples and a guide that describes how to quickly get a sample application up and running using a particular Steeltoe service. A detailed breakdown of the sample code is provided, describing how the Steeltoe service has been integrated into the app.
In many cases, these guides provide two ways of exercising the applications: one that describes how to create and run the application locally on your development machine, and a second that describes getting the application up and running on Cloud Foundry.
For the Quick Starts in which we run the application locally, we will at times make use of Java in order to run instances of the dependent servers (e.g. Spring Cloud Config Server, Netflix Eureka Server, etc.) locally on your machine. As such if you don’t have Java available on your machine you may want to install that now.
For the Quick Starts in which we make use of Cloud Foundry, you will need access to a Cloud Foundry environment that has the appropriate services (e.g. Spring Cloud Config Server, Netflix Eureka Server, etc.) installed. One option is to run PCF Dev, the local developer version of Pivotal Cloud Foundry on your development machine. PCF Dev makes use of Virtual Box and so depending on your desktop operating system and configuration you may or may not be able to make use of it.
Alternatively you can sign up for a free trial account of Pivotal Web Services the hosted multi-tenant edition of Pivotal Cloud Foundry. Note that if you want to work solely with .NET framework applications that target the Windows operating system, you’ll likely need access to a corporate Cloud Foundry environment, as neither of the above options currently support deploying Windows apps.
Regardless of which Cloud Foundry option you choose, in order to work with Cloud Foundry, you will need to install the Cloud Foundry Command Line Interface (CLI).
Finally, for all of the Quick Starts, you will need to install the GIT command line tools so you can fetch the Quick Start sample code and work with it on your computer.
Steeltoe has implemented a Eureka 1.0 client, not a 2.0 client. Eureka 2.0 is expected to have significant updates to its architecture and public API. At some point in time, we will examine a 2.0 implementation.
The Eureka client for Java contains features which enable operation on AWS. The Steeltoe version does not currently implement those features. Instead, this version has been optimized for Cloud Foundry environments. We may look at adding AWS cloud features at a future point in time.
Not all configuration properties found in the Java client are available for configuration. Those that are supported and provided by the Steeltoe implementation have been documented within.
Starting with version 1.1.0 of Steeltoe, the Steeltoe Eureka client allows you to configure what hostname/address gets registered with the Eureka server for your service registrations. This is provided by using a new setting,
spring:cloud:discovery:registrationMethod. By using this setting, you can now make use of Cloud Foundry Container to Container (C2C) networking support.
Unlike the Java version of the Config Server client, the Steeltoe client currently only supports property and yaml files, not plain text.
Steeltoe client only supports clear text communication with the configuration server, unless SSL/TLS is being used between the client and server. Client decryption is on our road map, but not currently supported.
Currently reloads must be initiated by the client, Steeltoe has not implemented handlers to listen for server change events. However this feature is expected to be added in the future.
Starting with version 1.1.0 of Steeltoe, you can now configure the timeout value the Steeltoe client uses when making Http requests of the Config Server.
Starting with version 1.1.0 of Steeltoe, the Config Server client is compatible with Spring Cloud and Spring Cloud Services Config Server deployments which support using Hashicorp Vault as back-ends.
Starting with version 1.0.1 of Steeltoe, the direct dependency on all of Oracle’s MySql packages have been removed. This has allowed the connector to become more flexible in terms of which MySql ADO.NET providers and versions it supports. With this release the connector is able to work with Oracle’s MySql Provider and the open source MySqlConnector providers.
For more detail on how to use these, see the MySql connector documentation. Note: This is a BREAKING change, as it now requires that you MUST explicitly include the packages and versions of the MySql/Entity framework code you wish to use in your application.
Starting with version 1.0.1 of Steeltoe, the direct dependency on the open source Npgsql package has been removed. This has allowed the connector to become more flexible in what PostgreSQL ADO.NET providers and versions it supports.
For more detail on how to use the connector see the PostgreSQL connector’s documentation. NOTE: This is a BREAKING change, as it now requires that you MUST explicitly include the packages and versions of the PostgreSQL/Entity framework code wish to use in your application.
Starting with version 1.0.1 of Steeltoe, the direct dependency on the open source RabbitMQ Client package has been removed. This has allowed the connector to become more flexible in what version of the client it supports.
For more detail on how to use the connector see the RabbitMQ connectors documentation. NOTE: This is a BREAKING change, as it now requires that you MUST explicitly include the packages and versions of the Rabbit code you wish to use in your application.
Also with this release, the connector has been updated to properly work with SSL/TLS based connections. Prior to this release, the connector did not support using the ‘amqps’ URL scheme.