Errors with Service Mapping- Missing pattern-based outgoing connections from load balancer. Please check the correctness of the incoming connection

vikrantsharma
Kilo Guru

We are performing Service Mapping on one of the Service We using, which we were able to Map on Istanbul version.

However, after we have upgraded to Kingston, we keep getting the error as below on the entry point itself:

 

Missing pattern-based outgoing connections from load balancer. Please check the correctness of the incoming connection

 

We are using the same setup with our MID Servers and other required configuration, and only thing which has changed in the Version of ServiceNow.

The Map is able to see the service running on the LB but it doesn't go anywhere from that. I have been looking for the resolution on the internet but couldn't find anything that is helpful.

We are on Kingston Patch 12. 

Please make it correct or helpful if this solves or help you with your issue for other to make use it.

Thanks & Regards,
Vikrant Sharma
9 REPLIES 9

UPDATING, I have found the solution!
That happened when you have F5 loader balancer. when the load balancer has no IRule, the pool wont forward the connection further, the Pattern will pick FQDN instead of the IP, so what you need to do:

  • Go your F5 in your CMDB -> load balancer pool 
  • Locate you application service and open it
  • Go down to load balancer pool members 
  • Under the name you find the ip addresss
  • Under the IP address you might find a fake FQDN instead of the ip ( here is the problem)  because under the ip address should be an ip matching the one under the name
 

HamzehAlsheikh_2-1732010583080.png

you can fix this manually, but whenever you run discovery, it will come up again as fqdn. to fix this problem permanently:

  • Go to Discovery Pattern -> search for  F5REST
  • F5 LB REST identification
  • Search for fqdn step

HamzehAlsheikh_3-1732010868111.png

change the fqdn attribute to ip

HamzehAlsheikh_4-1732011009147.png

  • Save -> push
  • Sync the pattern to the mid server -> run discovery again 

Please mark it green answer so everyone gets the solution 🙂 

HamzehAlsheikh
Tera Contributor

UPDATING, I have found the solution!
That happened when you have F5 loader balancer. when the load balancer has no IRule, the pool wont forward the connection further, the Pattern will pick FQDN instead of the IP, so what you need to do:

  • Go your F5 in your CMDB -> load balancer pool 
  • Locate you application service and open it
  • Go down to load balancer pool members 
  • Under the name you find the ip addresss
  • Under the IP address you might find a fake FQDN instead of the ip ( here is the problem)  because under the ip address should be an ip matching the one under the name
 

HamzehAlsheikh_1-1732011764507.png

 

you can fix this manually, but whenever you run discovery, it will come up again as fqdn. to fix this problem permanently:

  • Go to Discovery Pattern -> search for  F5REST
  • F5 LB REST identification
  • Search for fqdn step

HamzehAlsheikh_2-1732011764545.png

 

change the fqdn attribute to ip

HamzehAlsheikh_3-1732011763966.png

 

  • Save -> push
  • Sync the pattern to the mid server -> run discovery again 

Please mark it green answer so everyone gets the solution 🙂 

Community Alums
Not applicable

Hi @HamzehAlsheikh, For me its happening for Netscalar Load Balancer. I am not able to see any load balancer pool members on the service. Can you please guide me where I am going wrong.

Error: Missing pattern-based outgoing connections from load balancer. Please check the correctness of the incoming connection.

Hi @Community Alums , You need to run a horizontal discovery, when the ci (load balancer) is updated in the CMDB, click on the load balancer to open it, scroll down under the related links you should see the "Load balancer pools" tab. Open the tab and search for the application that you want to map. click on the application name to open it, there you can see under the related list "Load balancer pool members" where you see name and ip as well. the name and the ip should be identical. without that, the connection wont go further.

BUT what is more important is, after you done with horizontal discovery, you should go to the "discovery log" to know which pattern was kicked off. you can start debugging from there and find where it stopped.

Even when the step is green, there might be wrong in the table field mapping.
I hope the answer was helpful 🙂

Hi @Community Alums , I am facing same issue as yours. For me also its happening for Netscalar Load Balancer. Did you find any solution ? Please help. Thanks