Implemented wait-and-batch mechanism to group multiple alerts with the same fingerprint before sending a single notification, reducing alert fatigue and improving visibility.

**Alert Grouping/Batching:**
- Added `grouping.rs` module with in-memory batch storage using `DashMap`
- Implemented background worker in `alert_grouping.rs` polling every 1s for expired batches
- Integrated grouping logic in `scheduler/handlers.rs` after deduplication
- Supported all three `SendStrategy` variants: `FirstWithCount`, `Summary`, `All`
- Auto-send when `max_group_size` reached or timer expires after `group_wait_seconds`
- Registered background worker in `job/mod.rs`

**Observability - Prometheus Metrics:**
- Added 8 metrics to `metrics.rs`: dedup suppressions/passed/errors, grouping batches pending/sent/size/wait-time/errors
- Instrumented `deduplication.rs` to track suppressions and passed alerts by type (same-alert vs cross-alert)
- Instrumented `grouping.rs` and `alert_grouping.rs` to track batch lifecycle
- All metrics registered in Prometheus registry and exposed at `/metrics` endpoint

**UI Visibility:**
- Added dedup badges to alert names in `AlertList.vue` showing configuration status
- Added dedup column in `AlertHistory.vue` with visual indicators for sent/suppressed/grouped alerts
- Created `DedupSummaryCards.vue` component displaying org-wide stats (total alerts, dedup enabled count, suppression rate, pending batches)
- Added backend API `dedup_stats.rs` with `/alerts/dedup/summary` endpoint
- Removed legacy View History button from alert list page
- Extended `AlertHistoryEntry` with dedup fields: `dedup_enabled`, `dedup_suppressed`, `dedup_count`, `grouped`, `group_size`

**Logging & Debugging:**
- Comprehensive logging throughout grouping flow with `[grouping]` and `[alert_grouping_worker]` prefixes
- Enhanced deduplication logging with `[dedup]` prefix showing fingerprints and occurrence counts
- Added `get_pending_batch_count()` helper for API consumption

**Technical Details:**
- All features properly gated behind `#[cfg(feature = "enterprise")]`
- Backward compatible: grouping disabled by default
- In-memory batches cleared on restart (acceptable for 30s window)
- Thread-safe implementation using `DashMap` and atomic operations

Extended TriggerData with dedup fields for per-execution visibility and added compact dedup column to alert table.

Alert History Tracking:
- Added dedup fields to TriggerData: dedup_enabled, dedup_suppressed, dedup_count, grouped, group_size
- Implemented Default trait for clean initialization with ..Default::default()
- Set tracking fields in handlers.rs when alerts are suppressed, grouped, or sent
- History UI now shows actual dedup activity with icons

Alert List UI:
- Added compact Dedup column in alert table (80px width)
- Shows check icon if dedup enabled, dash if not
- Tooltip displays fingerprint fields and grouping config
- Clean inline status per alert without UI clutter

Fixed four failing unit tests:
1. test_organization_config_minimal_serialization - Updated to check correct field name 'alert_dedup_enabled' instead of 'cross_alert_dedup'
2. test_flatten_json_complex - Made JSON comparison order-independent by parsing embedded JSON strings
3. test_flatten_with_level - Added helper function to structurally compare JSON values regardless of key ordering
4. test_trigger_data_field_names - Updated field count from 21 to 26 to reflect new dedup/grouping fields

The tests were failing due to non-deterministic JSON key ordering and outdated field counts.

…groups

Implements org-level deduplication configuration with semantic field groups
for intelligent alert suppression and batched notifications.

Core capabilities:
- Semantic field groups: Map field name variations (`hostname`/`host`/`node`) to
  canonical dimensions for consistent deduplication across data sources
- Org-level dedup config: Global settings with cross-alert deduplication
  support to suppress alerts sharing semantic dimensions
- Alert grouping: Wait-and-collect batching with three send strategies
  (`FirstWithCount`, `Summary`, `All`)
- HTTP API: Endpoints for org-level deduplication configuration management

Features:
- Per-alert fingerprint-based deduplication with TTL expiration
- Cross-alert semantic matching using org-defined dimension groups
- Configurable time windows (per-alert override or org default)
- Background job processes expired batches every 1 second
- Three notification strategies for grouped alerts

Configuration:
- `SemanticFieldGroup`: Define field equivalences with optional normalization
- `GlobalDeduplicationConfig`: Org-level settings stored at `/alert_config/{org_id}/deduplication`
- `DeduplicationConfig`: Per-alert settings with fingerprint fields and grouping options
- Default presets for common semantic groups (host, IP, service, K8s resources)

API Endpoints:
- `GET/POST/DELETE /{org_id}/alerts/deduplication/config`

Implementation:
- Business logic: Pure algorithms in enterprise layer for fingerprinting and matching
- Service layer: Orchestrates DB operations with algorithm delegation
- HTTP handlers: Feature-gated dual implementations for OSS/enterprise builds
- Background jobs: Batch processor for grouped notification delivery

Extended TriggerData with dedup fields for per-execution visibility and added compact dedup column to alert table.

Alert History Tracking:
- Added dedup fields to TriggerData: dedup_enabled, dedup_suppressed, dedup_count, grouped, group_size
- Implemented Default trait for clean initialization with ..Default::default()
- Set tracking fields in handlers.rs when alerts are suppressed, grouped, or sent
- History UI now shows actual dedup activity with icons

Alert List UI:
- Added compact Dedup column in alert table (80px width)
- Shows check icon if dedup enabled, dash if not
- Tooltip displays fingerprint fields and grouping config
- Clean inline status per alert without UI clutter