Hi, I have an update.

This issue has been occurring since January, at least. I did another pass at an analysis of these crashes. I saw the last occurrence of An error occurred trying to start process '/agent/externals/node20_1/bin/node' with working directory '/agent/_work/1/s'. No such file or directory on the 3rd of March.

But, we started to see a different text of the issue. It happens in the same exact places and breaks pipelines the same way (described in the post above). The new error message is ##[error]StandardOut has not been redirected or the process hasn't started yet.. Not sure if that gives you a clue on what is wrong.

Our build agents are running on an image built on 17.2.2025. There hasn't been any change to them since. The version of all "problematic" pipeline tasks is the same as well.

Do you have any guidance?

I noticed this issue on a self-hosted agent today. On a whim, I simply copied the existing node20 folder to create a node20_1 folder. The pipeline then ran without issue. I don't know that I necessarily endorse this approach for a production system, but if you're self-hosted, and just need to move forward maybe give that a try as a workaround.

Here is the debug log of the failed script task:

2025-03-28T00:21:13.7232748Z ==============================================================================
2025-03-28T00:21:13.7232841Z Task         : Command line
2025-03-28T00:21:13.7232893Z Description  : Run a command line script using Bash on Linux and macOS and cmd.exe on Windows
2025-03-28T00:21:13.7233100Z Version      : 2.250.1
2025-03-28T00:21:13.7233159Z Author       : Microsoft Corporation
2025-03-28T00:21:13.7233224Z Help         : https://docs.microsoft.com/azure/devops/pipelines/tasks/utility/command-line
2025-03-28T00:21:13.7233307Z ==============================================================================
2025-03-28T00:21:13.8372243Z ##[debug]Using node path: /agent/externals/node20_1/bin/node
2025-03-28T00:21:13.8532074Z ##[error]StandardOut has not been redirected or the process hasn't started yet.
2025-03-28T00:21:13.8540765Z ##[debug]System.InvalidOperationException: StandardOut has not been redirected or the process hasn't started yet.
   at System.Diagnostics.Process.get_StandardOutput()
   at Microsoft.VisualStudio.Services.Agent.Util.ProcessInvoker.Dispose(Boolean disposing) in /mnt/vss/_work/1/s/src/Agent.Sdk/ProcessInvoker.cs:line 408
   at Microsoft.VisualStudio.Services.Agent.Util.ProcessInvoker.Dispose() in /mnt/vss/_work/1/s/src/Agent.Sdk/ProcessInvoker.cs:line 397
   at Microsoft.VisualStudio.Services.Agent.ProcessInvokerWrapper.Dispose(Boolean disposing) in /mnt/vss/_work/1/s/src/Microsoft.VisualStudio.Services.Agent/ProcessInvoker.cs:line 357
   at Microsoft.VisualStudio.Services.Agent.ProcessInvokerWrapper.Dispose() in /mnt/vss/_work/1/s/src/Microsoft.VisualStudio.Services.Agent/ProcessInvoker.cs:line 347
   at Microsoft.VisualStudio.Services.Agent.Worker.Handlers.DefaultStepHost.ExecuteAsync(String workingDirectory, String fileName, String arguments, IDictionary`2 environment, Boolean requireExitCodeZero, Encoding outputEncoding, Boolean killProcessOnCancel, Boolean inheritConsoleHandler, Boolean continueAfterCancelProcessTreeKillAttempt, TimeSpan sigintTimeout, TimeSpan sigtermTimeout, Boolean useGracefulShutdown, CancellationToken cancellationToken) in /mnt/vss/_work/1/s/src/Agent.Worker/Handlers/StepHost.cs:line 85
   at Microsoft.VisualStudio.Services.Agent.Worker.Handlers.NodeHandler.RunAsync() in /mnt/vss/_work/1/s/src/Agent.Worker/Handlers/NodeHandler.cs:line 267
   at Microsoft.VisualStudio.Services.Agent.Worker.TaskRunner.RunAsyncInternal() in /mnt/vss/_work/1/s/src/Agent.Worker/TaskRunner.cs:line 447
   at Microsoft.VisualStudio.Services.Agent.Worker.TaskRunner.RunAsync() in /mnt/vss/_work/1/s/src/Agent.Worker/TaskRunner.cs:line 76
   at Microsoft.VisualStudio.Services.Agent.Worker.StepsRunner.RunStepAsync(IStep step, CancellationToken jobCancellationToken) in /mnt/vss/_work/1/s/src/Agent.Worker/StepsRunner.cs:line 264
2025-03-28T00:21:13.8543273Z ##[section]Finishing: Node.js info

@tarunramsinghani Is there something we could try to narrow this down?

@ecl1ps, can you please help with what is agent version you are using ? and the OS and configuration(ram, storage etc) of the agent machine ? As it seems like something on OS is stopping the node process from getting started. I dont believe the cause is that the file is missing as mentoined in the initial comment as agent ships with node20_1.

Thanks for looking into this.

Allow me to answer your questions and also provide some additional details that might help with the investigation.

We’re using an Azure DevOps Agent Pool of type Azure Virtual Machine Scale Set.
Agent version is 4.252.0 (latest at the time of writing).

The VMSS behind the agent pool is configured as follows:
VM size: D2as_v4
RAM: 8 GB
vCPUs: 2
Local temp storage: 16 GB (SCSI)
OS disk: 128 GB Premium SSD (LRS), 500 IOPS
OS: Ubuntu 20.04 (sysprepped custom image from Azure Compute Gallery)

The issue appears intermittently. When I run a minimal YAML pipeline repeatedly on the same agent pool and image version, the problem does not reproduce. However, in actual production pipelines, the error does occur.

Given that the agent ships with node20_1, I also don’t believe the root cause is a missing file.
Could it be possible that something is modifying or misconfiguring the environment at runtime, causing the Node process to fail to start?

Any guidance or insight would be appreciated — let me know if further logs or debug output would help.

Thank you

Hmmm the disk space seems okay in this case. Would it be possible to collect and share diagnostics logs when this reproes again. To enable and collection diagnostic logs for agent please check this - Configure verbose logs

Here are logs of yesterday's failed run featuring StandardOut has not been redirected or the process hasn't started yet.
logs_1589538.zip

I see that AgentDiagnistic logs are missing, can you please enable the below option to make sure it will collect agent diagnostic logs.

Also seems like the Error "StandardOut has not been redirected or the process hasn't started yet." is masking the actual issue somewhere else. I still believe the original issue is the initial error you mentioned. Is it possible to repro this with 4.251.0 agent and with all diagnostic logs enabled and sysm.debug set to true as well please. that will be helpful in further debugging. Also can you please check if this reproes with any specific agent in the pool as I suspect there is some issue with FS which is causing the node command to fail. Maybe delete that specific VM if that is the case ?

The logs I posted above are everything that was generated when I enabled the system diagnostic by using

The Agent Diagnostics logs are just not there. I don't see a way to enable them other than system.debug. Or am I reading it wrong? https://learn.microsoft.com/en-us/azure/devops/pipelines/troubleshooting/review-logs?view=azure-devops&tabs=windows-agent#agent-diagnostic-logs

We are not using VMs but Scale Sets, I don't think the same applies to them.

For what its worth OPs issue isn't isolated. We have had the same error intermittently on agent version 4.252. OS is Windows for us

We're experiencing the same issue in our environment. We're also using Ubuntu-based agents running in a VM Scale Set. The error occurs consistently when the same Key Vault library is referenced in multiple jobs within the same pipeline.
Let me know if there's any recommended workaround or fix being worked on.

Although the Microsoft documentation doesn't list Node.js as a requirement, I found that installing it (Node version 24.2.0) on my self-hosted agent resolved the issue for me. Sharing here in case it helps others.