How can we achieve drift identification in redis cache?

Question

I have requirement where we will be using redis cache to read data for low latency which comes from persistent storage S3. So let’s say scenario where write operation happens to S3 (version 2) and it failed to write to cache because of some reason or gets corrupted, so now cache contains old data (version 1) only not the latest. So how can i achieve monitoring of data from s3 to redis cache, and make sure cache should always have the latest data. Thanks in advance.