Issue Summary

When setting cardinality Max >1 but not infinite, the attribute “expects” a single value, and removes older values, leaving only the last one.
Steps to reproduce:

Steps to Reproduce

1. Ensure you have an applink to Confluence
2. setup a Confluence Attribute on an object
3. Attribute config: set Card. Max to 5 - error message indicating 1 max
4. Set up Card.Max to infinite - accepted.
5. Create an object - select more than one page in the attribute - no problem!
6. Re-index of Assets (does not have to be clean)... after that, only the last Confluence page added is displayed!

Expected Results

It should continue to include multiple pages and display them.

Actual Results

Only the last Confluence page is stored in the attribute. There is no trace of the value removal in the Assets Audit log or the Object History.

The log shows behaviour similar to JSDSERVER-15551, although this was fixed in 10.2.0, and the current issue was reproduced on 10.3.13.

The logs show the re-index action (executed Re-index, not Clean Re-index!):

2026-01-29 14:14:07,953+0000 insight-InsightThreadGroup-worker-thread-4 WARN admin [c.r.j.p.i.s.core.impl.ClusterAwareReindexServiceImpl] Assets is locked for reindex 

2026-01-29 14:14:07,976+0000 insight-InsightThreadGroup-worker-thread-6 WARN admin [c.r.j.p.i.s.c.dal.impl.ObjectDalImpl] Found duplicate values for object with id 2, deleting them from the database 2026-01-29 

14:14:07,977+0000 insight-InsightThreadGroup-worker-thread-6 WARN admin [c.r.j.p.i.s.c.dal.impl.ObjectDalImpl] Duplicates resolved. Reloading objects from database - [2] 

2026-01-29 14:14:07,979+0000 insight-InsightThreadGroup-worker-thread-6 WARN admin [c.r.j.p.i.s.c.dal.impl.ObjectDalImpl] Sending cache update messages for objects - [2] 

2026-01-29 14:14:07,984+0000 insight-InsightThreadGroup-worker-thread-4 WARN admin [c.r.j.p.i.s.core.impl.ClusterAwareReindexServiceImpl] Assets is unlocked after reindex 

2026-01-29 14:14:07,985+0000 insight-InsightThreadGroup-worker-thread-4 WARN admin [c.r.j.p.i.s.core.impl.ClusterAwareReindexServiceImpl] Cluster lock released Assets clean reindex complete

Workaround

Currently, there is no known workaround for this behavior. A workaround will be added here when available

Versions Tested

Please confirm all versions that have been tested for this issue, and indicate whether the tested version is affected or not affected, below: