Hey, it does look like something is deleting the index. Did you have any stray DELETE requests? Based on the log timestamps I'm guessing there's something running a DELETE * on the cluster.
Can you share what kind of workload the cluster is under?