ML Based Service Mapping -- Application Fingerprint is not working even in Vancouver

[Anand D]

We moved to Vancouver last week. But Service Mapping Plus seems to be broken. In the Workspace the Application Readiness shows a message about AFP Training Job is Stuck. Also Most of the Source/Target AFP is missing for most of my maps. ML Powered candidates is empty.

I tried the troubleshooting guide via Troubleshooting guide for Service Mapping ML Connection Suggestions / Traffic Based Discovery - Supp... but didn't make a difference. Has anyone had similar issue ?

[AJ-TechTrek · Accepted solution]

Hi @Anand D ,

Refer the below Articles, which may help you.

https://www.servicenow.com/community/itom-forum/i-am-receiving-an-error-with-application-finger-prin...

https://docs.servicenow.com/bundle/washingtondc-it-operations-management/page/product/discovery/task...

https://support.servicenow.com/kb?id=kb_article_view&sysparm_article=KB0963421

Please appreciate the efforts of community contributors by marking appropriate response as Mark my Answer Helpful or Accept Solution this may help other community users to follow correct solution in future.

Thanks

AJ

Linkedin Profile:- https://www.linkedin.com/in/ajay-kumar-66a91385/

[Doug Mathias]

The Application Fingerprinting and Connection Suggestion Training jobs utilize Predictive Intelligence solutions to train data. If the solution has not completed successfully, these jobs will not be able to proceed and will be marked as "stuck".

Investigation Steps:

Open the ml_solution.LIST table from left Nav.

Add the "State" and "Update State" fields to the view.

In the "Table" field, search for *proc to filter ML solutions.

You should see two ML solutions: One for Application Fingerprints and One for Connection Suggestions

The ML solution is successfully completed if:

State = "Solution Complete"

Update State = "Update Complete"

If any of these fields indicate an error, it means the training has failed. 

In my situation, the ML Solution "ml_x_banhe_global_global_application_suggestion" had not been updated since 10/28/2023 and when we ran the AFP. the eventual outcome showed an "Error while updating the Solution" in the Update State field

I went through the various KB's that SN supplied but we had to open a SN Case as our AFP was repeatedly stalling. The case was escalated and SN devs got involved. It was found that "ml_capability_definition_clustering" was created in 2021, before a fix was released. Every new version was created from this record with wrong Capability, "Workflow Clustering", but should be "Clustering".

In the CASE I had opened, I was directed to: Remove this solution and create a new one using the steps from the link below. Basically the very bottom section "Recovery procedure for SM ML-based feature after trainer availability failure". This has to be done manually, not in a update set.

https://support.servicenow.com/kb?id=kb_article_view&sysparm_article=KB0963421