storage/src/main/java/org/zstack/storage/snapshot/group/VolumeSnapshotGroupChecker.java (1)

46-68: ⚡ Quick win

建议把逐组两次 count() 的查询改为批量聚合，降低热点路径开销

当前实现对每个快照组都发起两次计数查询，在 VM 组数量较多时会放大 DB 往返次数。建议一次性拉取 refs 后聚合（或使用分组统计），保持语义不变并减少拦截链路延迟。

可参考的重构方向

-        List<String> incomplete = new ArrayList<>();
-        for (Object o : groupUuids) {
-            String guuid = o.toString();
-            if (guuid.equals(excludeGroupUuid)) {
-                continue;
-            }
-            long deletedRefs = Q.New(VolumeSnapshotGroupRefVO.class)
-                    .eq(VolumeSnapshotGroupRefVO_.volumeSnapshotGroupUuid, guuid)
-                    .eq(VolumeSnapshotGroupRefVO_.snapshotDeleted, true).count();
-            if (deletedRefs == 0) {
-                continue;
-            }
-            long totalRefs = Q.New(VolumeSnapshotGroupRefVO.class)
-                    .eq(VolumeSnapshotGroupRefVO_.volumeSnapshotGroupUuid, guuid).count();
-            if (deletedRefs < totalRefs) {
-                incomplete.add(guuid);
-            }
-        }
-        return incomplete;
+        List<Tuple> refs = Q.New(VolumeSnapshotGroupRefVO.class)
+                .in(VolumeSnapshotGroupRefVO_.volumeSnapshotGroupUuid, groupUuids)
+                .select(VolumeSnapshotGroupRefVO_.volumeSnapshotGroupUuid, VolumeSnapshotGroupRefVO_.snapshotDeleted)
+                .listTuple();
+
+        Map<String, long[]> stats = new HashMap<>();
+        for (Tuple t : refs) {
+            String guuid = t.get(0, String.class);
+            boolean deleted = t.get(1, Boolean.class);
+            long[] s = stats.computeIfAbsent(guuid, k -> new long[2]); // [deleted, total]
+            if (deleted) {
+                s[0]++;
+            }
+            s[1]++;
+        }
+
+        return stats.entrySet().stream()
+                .filter(e -> !Objects.equals(e.getKey(), excludeGroupUuid))
+                .filter(e -> e.getValue()[0] > 0 && e.getValue()[0] < e.getValue()[1])
+                .map(Map.Entry::getKey)
+                .collect(Collectors.toList());

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In
`@storage/src/main/java/org/zstack/storage/snapshot/group/VolumeSnapshotGroupChecker.java`
around lines 46 - 68, The loop in VolumeSnapshotGroupChecker that calls two
per-group Q.count() queries on VolumeSnapshotGroupRefVO (one for
snapshotDeleted=true and one for total) is causing DB hot spots; replace it with
a single batched/grouped query that fetches counts per volumeSnapshotGroupUuid
(e.g., SELECT volumeSnapshotGroupUuid, SUM(snapshotDeleted), COUNT(*) GROUP BY
volumeSnapshotGroupUuid) for all groupUuids at once, then iterate the returned
aggregated results to populate the existing incomplete list (keeping the
excludeGroupUuid check and semantics of comparing deletedRefs < totalRefs)
instead of issuing per-group queries in the for-loop over groupUuids.

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes