"Streaming connection failed" while creating MLSkills for MLPackage on cloud uipath

“Streaming connection failed” while creating MLSkills for MLPackage on cloud uipath.
The provided URL “https://stream.launchdarkly.com/all” is not working

Anyone knows about this error?
Please help…

Hi @Pooja_Joshi27

Can you tell which ML skill you are trying and also can you provide the full error.

Also if you can go to ml logs there will be more detailed error description which might provide some insight.

It looks like the streaming url is not reachable , but need to know if it is for a specific ml or because of any issue on the package


Hi @Anil_G
I am trying to build out of box ML package, Where I have different formats of PDF files & from that I want extract data
I created dataset and trained it on cloud uipath. But while creating ML Skills getting error like this,


Hi @Pooja_Joshi27

I hope you are trying to deploy the minor version 1 or higher but not zero right?

and just to be sure. Can you run the evaluation pipeline as well and then try


Yes @Anil_G ,
I tried to deploy with minor version 1.
1st I created pipeline for package & when pipeline status become ‘successful’ then I deployed MLskills.

@sharon.palawandram any idea on this?

@Pooja_Joshi27 can you please gice the name of ml package…which out box ml it is?

Yes @Anil_G ,
ML Packages ->Out of the box Packages->UiPath Document Understanding->DocumentUnderstanding

  1. Created a new dataset. Then, data labeling. → Available.
  2. Created ML package then created pipeline for package → Successful
  3. Created MLskills for dataset → Available but log message contain “Streaming connection failed” error.

Hi @Pooja_Joshi27
I got the same issue, did you find any solution ?


Can you please explain what the URL is used for? “https://stream.launchdarkly.com/all”

Is it from your client side?

You might have to check the access and firewalls on the URL side as its failing the connection.

No @hussein.reda. Not found any solution.

I am encountering the same problem is there any solution?