Sometime in the last month and possibly after updating from 10.2 to 10.4 our full text indexing service has been chewing away at about 50% of the available CPU and memory.
The Queue is currently at 4.9 million documents and its processing about 30-50 documents a minute. Its going to take 3-5 months to complete.
I did a repository search for {LF:Indexed="N"} and it only found 390 documents that were not indexed. So it appears that nearly all of our ~5million documents are already indexed.
Anyone know what might have happened? How can we resolve it?
Could we have accidentally triggered a re-indexing? If so and we "clear the queue" will that stop the re-index?