Write-through caching writes every changed entry to both the cache and the underlying storage at the same time. The cache stays available for fast reads, while persistent storage is updated immediately. section 7.6
Use the policy when the workload favors reliable, repeated reads over faster writes:
- Choose it for read-intensive applications where data is written once or infrequently and then read many times. section 7.6
- Expect slower writes than with write-behind caching, because each change reaches the main storage immediately. section 7.6
- Reduce data-loss risk to nearly zero except for edge cases and malfunctions, while retaining caching benefits for reads. section 7.6
Session data
A session-data system can write each update to the cache and database together, then serve subsequent reads from the cache. This suits data written once or seldom and read repeatedly soon afterward. section 7.6
Unlike write-behind, which delays storage writes, write-through does not improve writing performance. Unlike write-around, it does not leave newly written data out of the cache. Read-through instead describes how entries are populated when read, not this simultaneous write policy. section 7.6
