Inconsistent search results in different threads but same query [LUCENE-5566]

We have a very difficult case to identify whats going on.
We have a multithreaded environment where multiple threads are trying to hit the same lucene query and getting different results each time.
For eg: Query A has returned 1 result which is running in thread 1
And same Query A is returning 0 results which is running in thread 2.

All the indexes looks fine and if we take the same query and try it in Luke and we are able to get the correct results which is 1 in this case. And we are not able to reproduce it in local environments or staging or integration but only on Production.

Trying to figure out whats going on.

---
Migrated from [LUCENE-5566](https://issues.apache.org/jira/browse/LUCENE-5566) by Pradeep K Poluri

Start by comparing the multithreaded production execution described in the issue with running the same query in Luke, where the expected result is returned. Investigate why identical queries can produce different results across threads and determine what additional production conditions or diagnostics are needed; done means the cause is identified and a reproducible fix or clear explanation is available.

Written by the indexing model from the issue text.