microsoft / microsoft/onnxruntime

Improving requirements and dependency check for deploying

Open
#14,064 13 comments 1 reaction 0 assignees View on GitHub

Nobody has claimed this yet.

documentation ep:CUDA ep:DML platform:windows
Dominant language
C++
Stars
21.9k
Forks
4.2k
Avg merge
4d 8h
Merged PRs (30d)
179

Description

Describe the issue

We're working on a product using ORT that is to be deployed to clients in the wild, as discussed in #13276, when deploying in the wild it's not possible to know in advance what's installed or not in the machine before running.

I am aware that DirectML and CPU providers require a certain version of VCRedist to be installed in the machine in order to run.

I need to add the VCRedist version check to my installer, so if the VC redist is missing, prompt the user to install the required runtimes... So I need to know which is the exact minimum version required by the providers to run.

I looked around other issues, pointing to https://github.com/Microsoft/onnxruntime#system-requirements but that link is broken.

To reproduce

Just try to use OnnxRuntime in a brand new machine, 5 minutes after installing Windows OS, and you'll see how many things fail...

As it happens, most of our clients are buying new machines just to use our software, and most of them are not tech savvy, So we end up having to guide them manually, by phone, to check the errors and install the proper runtimes (Window updates up to the right version, VC Redist, Graphics drivers, CUDA, etc). You don't realize how painful it is until you suffer it.

Urgency

I am writing the installer right now, so at the very least, I would need to know the VC Redist minimum version used by ONNX

Platform

Windows

OS Version

Win10

ONNX Runtime Installation

Released Package

ONNX Runtime Version or Commit ID

1.13.1

ONNX Runtime API

C#

Architecture

X64

Execution Provider

DirectML

Execution Provider Library Version

1.13.1

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reviewing the deployment and system-requirements material referenced in the issue, including the broken system-requirements link and discussion in #13276. Determine whether the change belongs in ONNX Runtime's released package documentation or in an installer, then define the supported DirectML and CPU prerequisites and how completion will be verified.

Written by the indexing model from the issue text.

Assessment

Tech stack
csharp
Domain
devops, operating-systems
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.