Have you experienced significant differences in the volume of data stored in shards at your previous job?

Yes, in data sharding projects, there are often significant differences in data volume between shards. This can be due to uneven key distribution or load characteristics. For example, if sharding is based on hashing by user key, but some users generate much more data than others, the corresponding shards will contain more information. To address this issue, data rebalancing, changing the sharding strategy, or using more complex distribution algorithms are employed to ensure a more even distribution of load and data volume.