Review Performance

Performance Review

Procedure

Step 1 — Check for N+1 queries

• Search for loops that call the database:

  grep -rn "\.findById\|\.findAll\|\.existsById" openaev-api/src/main/java/ --include="*.java"
  

• Verify that no findById() or findAll() is called inside a for / forEach / stream().map()
• If a loop needs related entities, prefer findAllById() or a single @Query with IN clause
• Check @ManyToMany / @OneToMany collections have @Fetch(FetchMode.SUBSELECT) to avoid N+1

Step 2 — Check fetch strategy

• All associations should default to FetchType.LAZY
• FetchType.EAGER is only acceptable for small, always-needed collections (e.g. capabilities)
• Search for EAGER on potentially large collections:

  grep -rn "FetchType.EAGER" openaev-model/src/main/java/ --include="*.java"
  

• Verify that LAZY collections are never accessed outside a transaction (causes LazyInitializationException)
• For API endpoints returning IDs only: LAZY + subselect is preferred

Step 3 — Check pagination

• All list/search REST endpoints MUST return Page<T>, not unbounded List<T>
• Search for endpoints returning lists:

  grep -rn "List<.*>" openaev-api/src/main/java/io/openaev/api/ --include="*.java" | grep -i "public\|return"
  

• Verify reasonable default page size (10-20) and max page size (100)
• findAll() without pagination is only acceptable for small reference data tables

Step 4 — Check query efficiency

• Search for findAll() that could be filtered at DB level:

  grep -rn "\.findAll()" openaev-api/src/main/java/ --include="*.java"
  

• Verify existence checks use existsById() instead of findById().isPresent()
• Verify bulk deletes use @Modifying @Query instead of loading + deleting one by one
• Check that @Transactional(readOnly = true) is used on all read methods

Step 5 — Check database indexing

• New columns used in WHERE / ORDER BY / JOIN should have indexes
• FK columns in join tables should be indexed (composite PK covers one direction, check the other)
• For new migrations, verify:

  grep -rn "CREATE TABLE\|CREATE INDEX\|ADD COLUMN" openaev-api/src/main/java/io/openaev/migration/ --include="*.java"
  

Step 6 — Check memory usage

• No large byte[] or full file content loaded in memory — use streaming
• No unbounded in-memory collections (e.g. findAll() result stored in a List)
• Search for potential memory issues:

  grep -rn "byte\[\]\|ByteArrayOutputStream\|toByteArray" openaev-api/src/main/java/ --include="*.java"
  

Step 7 — Check transaction scope

• Read-only operations use @Transactional(readOnly = true)
• No long-running computation inside @Transactional (keep transactions short)
• Check for MinIO / S3 / file I/O inside @Transactional: search for calls to fileService or uploadCatalogLogo inside methods annotated @Transactional — flag as blocking issue (DB connection held during network I/O). Required fix: split into initialise() (DB, transactional)
  • refreshAssets() (MinIO, non-transactional) called in sequence after the transaction commits.
• Verify no @Transactional self-calls (Spring proxy bypass):

  grep -rn "this\." openaev-api/src/main/java/io/openaev/service/ --include="*.java" | grep -i "find\|get\|search\|create\|update\|delete"
  

Step 8 — Report

Document findings using conventional comments format:

• issue (blocking): for performance bugs (N+1, missing pagination, unbounded queries)
• suggestion (non-blocking): for optimizations (index, fetch strategy, caching)
• note: for informational items (acceptable tradeoffs, future improvements)