Wazuh alert manager v1.5.1
v1.5.1
A small patch release on top of v1.5.0: it fixes a correctness race in single-tick burst correlation and ships a reliability benchmark / reproducibility kit. It contains everything in v1.5.0 (alert lifecycle, cases, automation rules, precedent & suggested alerts, reporting).
Fixed
- Single-tick burst correlation no longer misses. When an entire alert burst arrived within one sync interval, the automation engine could evaluate it before those alerts were search-refreshed in wazuh-alert-status, so the count raced the ~1 s refresh and a burst could be silently missed (and, because the alerts leave the next tick's window, never re-evaluated). The sync now refreshes the status index before running rule evaluation, making single-tick burst detection deterministic. (server/lib/sync_job.ts)
Added
- Reliability benchmark / reproducibility kit (benchmark/). Injects a tagged 1,082-alert synthetic corpus with known-correct outcomes through the real sync → automation path and measures the engine's actions, including false actions on benign filler alerts. See benchmark/README.md.
Verification
- Regression smoke: 16/16 passing on Wazuh 4.14.
- Reliability run: 1,082 / 1,082 alerts handled correctly, 0 false actions on 1,000 benign fillers.
Downloads (match the asset to your Wazuh version)
| Wazuh | OpenSearch Dashboards | Asset |
|---|---|---|
| 4.12 | 2.19.1 | wazuhAlertManager-1.5.1-wazuh4.12.zip |
| 4.13 | 2.19.2 | wazuhAlertManager-1.5.1-wazuh4.13.zip |
| 4.14 | 2.19.5 | wazuhAlertManager-1.5.1-wazuh4.14.zip |
Server code is identical across the three builds; only the target-version stamp
differs. Install/upgrade steps are unchanged — see the
wiki.
What's Changed
- Add citation metadata (CITATION.cff, .zenodo.json) and license field by @xrisbarney in #5
- Add Zenodo DOI badge by @xrisbarney in #6
- fixes to 1.5.1 by @xrisbarney in #7
Full Changelog: v1.5.0...v1.5.1