In my previous post, I showed how I collect index information across all replicas to get the whole picture. In this follow-up, I add more context.