OData / odata.net

ODataLib: Open Data Protocol - .NET Libraries and Frameworks
https://docs.microsoft.com/odata
Other
687 stars 349 forks source link
# OData .NET Libraries Build Status
Rolling
Nightly

1. Introduction

The OData .NET Libraries (or OData .NET, for short) project includes the implementation of core functionalities of OData protocol on the .NET platform, including URI parsing, request and response reading and writing, Entity Data Model (EDM) building, and a .Net OData client which can be used to consume an OData service. It is a fully open sourced project maintained by Microsoft OData team. The libraries are used by the WebApi and RESTier libraries for building OData Services.

OData stands for the Open Data Protocol. It was initiated by Microsoft and is now an ISO ratified OASIS standard. OData enables the creation and consumption of REST APIs, which allow resources, identified using URLs and defined in a data model, to be published and edited by Web clients using simple HTTP requests.

For more information about OData, please refer to the following resources:

For how to adopt this and related libraries to build or consume OData service, please refer to the following resources:

2. Project structure

The project currently has a main branch and three archived branches: maintenance-6.x, maintenance-5.x, and maintenance-wcf-dataservice-v4.

main branch:

The main branch is the active development branch for ODataV4 7.x. It produces libraries targeting .NET 4.5 as well as .NET Standard 1.1 and 2.0. The branch builds with Visual Studio 2019.

For each profile above, it has the following libraries:

For these libraries, we accept issue reports and welcome contributions through pull requests.

dev-8.x branch:

The dev-8.x branch is the active development branch for ODataV4 8.x.

maintenance-6.x branch: (maintenance mode)

The maintenance-6.x branch includes the .NET libraries for ODataV4 6.x maintenance releases.

maintenance-5.x branch: (maintenance mode)

The maintenance-5.x branch includes the .NET libraries for OData V1-3 releases only. It has the similar libraries as the maintenance-6.x branch except for some differences in namespaces and two additional libraries:

These libraries are in maintenance mode. Only security bugs will be accepted. The release will be irregular depends on the bugs fixed.

maintenance-wcf-dataservice-v4 branch: (maintenance mode)

The maintenance-wcf-dataservice-v4 branch has the source code of the OData V4 parity of the WCF Data Services Server for OData v1-3. It is provide for demo purposes only; it has no binary release and does not accept contributions. WebApi or RESTier is recommended for building new OData Services.

3. Building, Testing, Debugging and Release

Since we are building this on VS2019 LocalDB v12.0 or above will be used which is part of VS2019 and no additional installation is needed. The Database will be automatically initialized by the test code if it doesn't exist.

Note: The project T4CrossPlatformTests.WindowsStore.csproj will not be loaded unless you have installed the Windows 8.1 and Windows Phone 8.0 / 8.1 tools.

3.1 Building and Testing in Visual Studio

Simply open the shortcut OData.sln at the root level folder to launch a solution that contains the product source and relevant unit tests. Should you see the need to modify or add additional tests, please see the sln folder for the whole set of solution files.

Here is the usage of each solution file (the OData.sln shortcut opens the one marked default):

Each solution contains some test projects. Please open it, build it and run all the tests in the test explorer. For running tests within OData.E2E.sln you need to open Visual Studio IDE as Administrator so that the test services can be started properly.

3.2 One-click build and test script in command line

Open Command Line Window with "Run as administrator", cd to the root folder and run following command:

build.cmd

This will build the full product and run all tests. It will take about 60 minutes. Use the to ensure your change compiles and passes tests before submitting a pull request.

Optionally, you can run following command:

build.cmd quick

This will build a single set of product Dlls and run unit tests. It will take about 5 minutes. Use this for quickly testing a change.

Here are some other usages or build.cmd:

Notes: If your receive build error with message "build.ps1 cannot be loaded", right click "build.ps1" -> Properties -> "Unlock".

3.3 Debug

Please refer to the How to debug.

3.4 Nightly Builds

The nightly build process will upload NuGet packages for ODataLib (Core, Edm, Spatial, Client) to the MyGet.org odlnightly feed.

To connect to odlnightly feed, use this feed URL: odlnightly MyGet feed URL.

You can query the latest nightly NuGet packages using this query: MAGIC OData query

3.5 Official Release

The release of the component binaries is carried out regularly through Nuget. A new version is released every 2 months. A new milestone item will be created after each release. It will correspond to the work that is expected to be included in the next release. Any work that has been completed by the due date for the milestone will be shipped regardless of if the entire milestone is completed. Work that is not completed by the due date will be moved to another milestone.

Details on the release process are found here

3.6 Performance benchmarks

Installation

The easiest way to run the perf benchmarks is to use the Microsoft.Crank toolset.

Start the agent

Run benchmarks locally

Run benchmarks on remote dedicated agents

The local profile is provided for testing purposes, but it's not ideal for running benchmarks. For more stable results and results that we can more reliably compare, the following profiles are also available and should be preferred whenever possible:

Profile Machine Architecture OS
lab-windows INTEL, 12 Cores Windows Server 2016
lab-linux INTEL, 12 Cores Ubuntu 18.04, Kernel 4.x

Use the --profile argument to specify the profile you want to use. For example, to run the components benchmark on the Windows agent, run the following command:

crank --config benchmarks.yml --scenario Components --profile lab-windows

And to run on the Linux agent:

crank --config benchmarks.yml --scenario Components --profile lab-linux

PS: We should not use these machines to run automated scheduled benchmarks.

Run benchmarks against the official repo

To run benchmarks against the official repo instead of your local repo, pass the base=true variable to the command, e.g.:

crank --config benchmarks.yml --scenario ODataWriter --profile local --variable base=true

This will cause the crank agent to clone the official repo and run the tests against the main branch.

You can specify a different branch, commit or tag using the baseBranch variable:

crank --config benchmarks.yml --scenario ODataWriter --profile local --variable base=true --variable baseBranch=v7.6.4

Run load tests

Besides benchmarks, we also have some load tests which measure request round-trips from a client to a server. These can be used to evaluate how OData libraries behave when handling multiple concurrent requests on the same server.

We have tests that evaluate different writer implementations, serializing a simple static collection response on each request:

crank --config loadtests.yml --config lab-windows --scenario SerializationComparisons --application.options.counterProviders System.Runtime --variable writer=ODataMessageWriter

This scenario allows you to choose which writer implementation is used to process the response as well. It also allows you to configure different aspects of the requests, e.g. number of connections, max requests per second, etc.

For more information about these tests, read this doc.

Collecting traces

Crank can collect and download native trace files from the benchmarked application that you can analyze in specialized tools using the --[job].collect true option, where [job] is the name of a job defined in the .yml config file.

On Windows, --[job].collect true option will collect traces using PerfView download an .etl trace file that you can you can also analyze using PerfView.

For example, the loadtest.yml config defines an application job which refers to the server handling the requests. We can collect traces from the server as follows:

crank --config loadtests.yml --config lab-windows --scenario SerializationComparisons --variable writer=ODataMessageWriter --application.collect true

Comparing benchmarks

You can use the ResultsComparer tool to compare benchmark results.

Example:

cd tools/perf/ResultsComparer/src/ResultsComparer

dotnet run -- --base=BenchmarkBefore.json --diff=BenchmarkAfter.json --threshold 1%

Learn more about the benchmark results comparer here.

4. Documentation

Please visit the ODataLib docs on docs.microsoft.com. It has detailed descriptions on each feature provided by OData lib, how to use the OData .Net Client to consume OData service etc.

5. Community

5.1 Contribution

There are many ways for you to contribute to OData .NET. The easiest way is to participate in discussion of features and issues. You can also contribute by sending pull requests of features or bug fixes to us. Contribution to the documentations is also highly welcomed. Please refer to the CONTRIBUTING.md for more details.

5.2 Reporting Security Issues

Security issues and bugs should be reported privately, via email, to the Microsoft Security Response Center (MSRC) secure@microsoft.com. You should receive a response within 24 hours. If for some reason you do not, please follow up via email to ensure we received your original message. Further information, including the MSRC PGP key, can be found in the Security TechCenter. You can also find these instructions in this repo's SECURITY.md.

5.3 Support

Thank you

We’re using NDepend to analyze and increase code quality.

NDepend

Code of Conduct

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.