I have a experiment stuck in qued status

indrani sen 1

I have a time series experiment which is indefinitely stuck on queued status. It was successfully running before . But I added a new module for installing a new R library from repository and following that everything is queued.
I created a new copy of the experiment and deleted the old one but still it seems to be running. Please help

Ramr-msft 17,616 Reputation points

2021-08-20T15:26:11.457+00:00

@indrani sen Thanks for the question. Could you please share the error details to debug further on this.
indrani sen 1 Reputation point

2021-08-20T18:51:03.983+00:00

This is my workspace.
https://studio.azureml.net/Home/ViewWorkspaceCached/aa1e2a13c03a4deea657faaae8c48d01?#Workspaces/Experiments/Experiment/aa1e2a13c03a4deea657faaae8c48d01.f-id.def9055907074f6d86446fbe242a7df0/ViewExperiment

(In the running status it shows "your experiment has been submitted and is in the queue to be run") in the run history all the experiments are locked for editing.

No experiments are running now. All are getting queued. I deleted the erroneous experiment.
I had to create a new account to create a new editable experiment.

There was no error. I was trying to install a R library into the ML studio. The experiment got queued forever.
There is no way to terminate a infinite running process here.

This is my code

uri_repo <- "file:///C:/src/myrepos/"
options(repos = uri_repo)
inputlib <- "C:/src/myrepos/"

mypackages <- c("timetk")
outputlib <- "serverrepo"

install.packages(mypackages, repos = file.path("file://", normalizePath(inputlib, winslash = "/")), type = "win.binary", dependencies = TRUE)

While running the last line(install.packages) the experiment gets queued. The lines before gets executed.

There may be errors but the whole workspace is affected because of this. This is a serious bug. @Ramr-msft

1 answer

Ramr-msft 17,616 Reputation points

2021-08-24T05:22:29.783+00:00

@indrani sen Thanks, We would recommend to provide a feedback item along with screen shot of this error to the service team from Azure ML portal to help you on this. Navigate to the page of your run, click the smiley icon on top right corner and provide details of the issue and select "include screenshot" option along with option to contact you for feedback. This should enable the team to complete your request and reach out to you for more information.
Please sign in to rate this answer.

0 comments No comments
Sign in to comment