A small marketing company recently moved all of its storage to a cloud provider. The budget for storage has limited the company to buying a small amount of SSD storage and a lot of spinning disks…

Compression reduces the storage footprint of data but does not improve read/write speed and can actually add CPU overhead that degrades performance.

Automated storage tiering uses policies or I/O monitoring to dynamically migrate data between storage tiers - placing hot, frequently read data on SSD for maximum throughput and latency performance, while keeping cold data on cheaper spinning disks. This approach fully leverages the limited SSD capacity the company purchased while still keeping costs low, achieving the best possible speed within the given budget.

Deduplication eliminates redundant copies of data to save space but does not accelerate I/O or direct data to faster storage media.

Autoscaling adds or removes storage capacity dynamically but does not optimize the performance of existing storage by prioritizing fast media for hot data.