Hibernate / MyBatis Interview questions
What is the difference between first-level and second-level cache?
Both caches avoid redundant database queries, but they differ substantially in scope, default status, and the kind of staleness risk they carry.
| First-Level Cache | Second-Level Cache |
| Scoped to a single SqlSession. | Scoped to a mapper namespace, shared across sessions. |
| Enabled by default, cannot be turned off. | Disabled by default; must be explicitly enabled. |
| Cleared automatically on commit/rollback/close. | Persists across sessions until explicitly flushed or evicted. |
| Low staleness risk, given its short-lived scope. | Higher staleness risk if misapplied to frequently-changing data. |
The practical implication is that the first-level cache can be relied on essentially without thought — it's automatic and safely scoped — while the second-level cache is a deliberate performance optimization that requires understanding a specific mapper's data access patterns well enough to judge whether cross-session caching is actually safe for that particular data, rather than something to enable indiscriminately across every mapper.
Which cache can be shared across multiple different SqlSession instances?
Which cache carries higher staleness risk if misapplied?
More Related questions...
