Orchestrator Process "...Input Parameters..." missing when using external NuGet server

I wonder if someone can explain what I’m doing wrong, or whats not correctly configure, or if this is a potential product fault.

I am trying to use the source code example provided at:

When i create a process in Orchastrator I get them message “The current package version has no input or output parameters.”

Orchestrator has been configured (through the web page) to use an external NuGet server. Specifically, Settings > Deployment: Package deployment is set to a external url (aka, a different server in my network that has Internet access). Libraries is set to Tenant, External url (the same server as above but different repos).

Both the NuGet urls are the same server but different repos. (Using Nexus 3 NuGet server). Both are pointing to groups. The package group contains only locally hosted repos; this is configured in studio as is publishing location. The Library group contains a locally hosted repos also configured in studio AS WELL AS proxy definitions to the standard UiPath internet repos’s.

Using Orchastrator web gui, all is fine, I can see things I am publishing, things I need. All robots, studios are linked and can see these locations.

HOWEVER, for the UiPath provided example process, the WebGui never gives you the opportunity to set inbound parameters. You only get the message “The current package version has no input or output parameters.”. This is also true for any packages I write and publish.

QUESTION: Whats wrong with my setup, what have I not understood to make this work?

1 Like

Just as a thought. When does the package get downloaded? At execution time or at process configuration time?

Is this a case that setting up the process in Orchastrator uses the package name but the actual files for that package are not downloaded until later. ie. Orchastrator is not aware of what the input parameters are since it can’t open the package file locally.

hi @powersg
is this problem solved already or do the input parameters still don’t work.

Other question: what will happen when Uipath Orchestrator gets updated? Where are the new core activities placed then? Still have to test this, but maybe you already have.


i wil test this further. I have 2 different (Test) instances of Uipath Orchestrator. On 2019.10.14 i get no input parameters, on 2019.10.16 i now get the input parameters when i connect it to Nexus3.

Will do some advanced testing on this. Will update this post if i have any new info.

This problem i still have to test: Impact on upgrade of Uipath. what will that do to core Libraries

1 Like

Cool … It’s not just me then!!!
I have, at this time, hard-coded the input parameters for the things I needed or switched to using a local xml cfg file. I am however getting close to this problem having to re-investigated from my point of view whats going on.

If Nexus is supported as I would expect on 2019.10.16++, sounds like I need an upgrade. I’ve been tracking LTS versions at this time to minimise impacts. I’m still on 2019.10.1

Keep us informed! Very keen to get this working! Would make my life easier.

Thx for the reply!.
I have an “old” uipath trial on 2019.10.14. It is currently unlicensed, but for testing this setup that doesn’t matter. I got a second setup on Server 2019 which is 2019.10.16
But i don’t know when this will be released as an LTS version. Maybe / hopefully in the April release.

I agree on making life easier. Especially when the orchestrator / robot environments don’t have internet access to download the packages from the nuget/myget repositories.

Connecting everything to Nexus would make everything easier. You can already publish from the Studio to Nexus repository, so nexus can become the central place for version control of the package/library releases.

But like i said, i’m curious what an upgrade would do to the core packages. Only one way to find out i guess. Keep you posted.

My Robots/Orchestrator do not have internet access due to their location within my network topology. My nexus server does, however have internet access (be it restricted). I therefore have setup nexus proxy caches to point to the standard UiPath repo’s (i.e., those defined by UiPath by default). This seems to be the way to go.

Thx. Will look into that. Not that familiar with nexus yet.

I am having the same issue, my Non-prod Orchestrator and Prod orchestrator are pointing to the same external nexus package repo. My Non-prod Orchestrator (Which developers push to and feeds into the external nexus repo) properly shows input arguments to packages when configuring the automations in Non-Prod Orchestrator. However, my production Orchestrator does not populate package inputs for the same exact processes and versions (looking at the same package repo). However if I change my production orchestrator to Internal repo and manually upload the package, the automations have input arguments. If the problem is that orchestrator needs to download the package from the external repo to display arguments properly, how can I do this?

Hi @Noppy , Which UiPath versions are you Orchestrator machines? What if you point your Non-prod to Nexus repo1 and your prod-Orchestrator to nexus repo2. Does this solve the issue?

I would not point both machines to the same repo (you don’t have your non-prod and prod running on the same machine either i assume).

Will see if i can find some time to do the setup again with a Nexus non-prod and prod orchestrator. I don’t have running On-premise machines at the moment, so i will see to setup a trial setup soon.

Thank you for the reply. After speaking to technical assets at UiPath we figured out that Orchestrator will not get the necessary metadata for a process by only reading from a repo. To have parameters show up in Orchestrator the process artifact must be published directly to Orchestrator (which Orchestrator can then update the downstream repo). Publishing could occur through studio or other CI/CD pipelines. The publishing event is what creates the necessary metadata information in Orchstrator DB to load the parameter options.

At this time we will be manually importing artifacts to Prod Orchestrator until we have a pipeline to do so through Gtilab or Jenkins.

Hi, I am facing the same issue, any fix for this?