The post discusses the debate around the necessity of Redis in certain architectures, emphasizing that the choice between using Redis and traditional databases depends on specific use cases, particularly those involving high velocity data updates. It highlights that while Redis excels in scenarios involving counters, news feeds, and chat messages due to its efficiency in managing high update rates with less overhead, smaller scale operations (like handling under 10,000 queries per second) may not benefit from its use and can perform adequately with simpler solutions such as MySQL. The conversation also touches upon the key considerations in caching strategies, including understanding when to use cache, ensuring fallbacks for cache invalidation, and the potential drawbacks of premature optimization. Additionally, opinions vary on Redis, with some deeming it essential while others criticize it, indicating a broad spectrum of preferences in tech stacks, even leading some companies to completely avoid it in favor of alternatives.