- Applies to 8.24.7 and earlier.
- Applies to clusters created during Check Point normalization.
If there are already two or more cluster members named the same, normalization will through the following error after detecting the duplicate cluster names:
2019-06-12 19:18:38,400 [NIO-2 task-125] [ERROR] c.f.p.a.jaxrs.handler.ExceptionHandler : Logged at the oubound exception handler org.springframework.dao.IncorrectResultSizeDataAccessException: result returns more than one elements; nested exception is javax.persistence.NonUniqueResultException: result returns more than one elements
Take a look under Administration --> Device --> Clusters and sort by name. You may find there to be an excessively large number of cluster objects.
If so, you'll need to use the powershell script that is attached to this ticket. You will need to run this from a PC that has Powershell and access to the App server. Rename the script to delete_cluster_objects.ps1 and then run it in Powershell ISE. Run by right clicking on the file and select run as Powershell ISE.
The script will remove all empty cluster objects. Once done, you'll need to go Administration -> Device --> Clusters
and you will be left with duplicate cluster objects where each cluster object will have one member. You'll need to delete one of the clusters and then add the member to the other remaining cluster object. There's unfortunately no easy way to do this.
Shut down the Data Collector(s) while doing this so that new revisions do not normalize and cause more clusters to be generated.
Log into the data collector via CLI and run:
fmos maintenance begin
If they do, you can always re-run the script to clear out the empty ones.
The script needs to be renamed to .ps1 and opened in "powershell ISE" and run from there. It will prompt them for their app server IP, username and password and then delete empty clusters.
If Windows returns 'Permission Denied,' open Powershell (not powershell ISE) as administrator and run:
Once completed, log back into any Data Collectors from the CLI and re-enable them:
fmos maintenance end
Article is closed for comments.