How to Install and Explore Nightingale v7.7: New Features, Upgrade Guide, and Hands‑On Demo

Nightingale Monitoring released version v7.7, the last release of the v7 series, offering minor fixes and an improved user experience; development of v8 will begin next week with a rewritten notification logic and support for ElasticSearch, ClickHouse, and machine‑alert policies linked to business groups.

v7.7 Main Changes

feat: alarm rule data source filtering, supports inverse selection and fuzzy matching

feat: alarm rule query condition can set "unit"

feat: alarm rule supports "override global callback" setting

feat: alarm rule Prometheus source preview adds Step setting

refactor: dashboard detail page adds link to return to dashboard list

refactor: tdengine data source compatible with v2 version

fix: resolve Elasticsearch source legend template issue with variable names containing "."

fix: edge module machine loss alarm tag loss issue

doc: improve alarm level name terminology

doc: add Doris dashboard and alert templates

Important v7 Major Changes

Global dark theme

New metric view with hundreds of built‑in promql queries

Template center for creating and managing templates

Machines can be bound to multiple business groups

Optimized edge‑site machine loss alarm logic for closed‑loop handling

Notification tag filtering to avoid irrelevant tags

Global callback address page enhanced with detailed documentation

Support sending alerts to DingTalk, Feishu, WeCom via callback URLs

Built‑in self‑healing fault capability without separate ibex module

Dashboard variables linked to machines of the current business group

Machine list and metric view integrated for multi‑machine charting

Alarm rules can configure PromQL for recovery notifications

Integration of Grafana dashboards into Nightingale

Upgrade Method

Upgrade a v7 minor version by replacing the binary and

integrations

directory, or pull the latest Docker image. Because v7.7 changes the edge module, also upgrade

n9e-edge

if it is used.

Database schema changes are centralized; if your DB account cannot modify tables, execute the provided

migrate.sql

manually.

Project Introduction

Nightingale Monitoring was open‑sourced by Didi in 2020, later donated to the China Computer Federation and managed by its Open‑Source Development Committee. It focuses on monitoring, supports multiple data sources with a unified UI for alert rules and notifications, and includes built‑in visualization and dashboard templates.

Installation Test

Use Docker to test the installation: download the v7.7 release package from https://flashcat.cloud/download/nightingale/ , unzip, navigate to

docker/compose-bridge

, and run:

<code>docker compose up -d</code>

Note: if you have previously pulled the Nightingale image, pull it again to get the updated version. After the containers start, open a browser to port 17000 to access the UI, log in with root / root.2020 , and change the password.

Because Nightingale supports multiple data sources, add a data source via the Integration Center → Data Sources (default is VictoriaMetrics):

Click Add , choose a Prometheus‑like data source, fill in the name and address (e.g.,

http://victoriametrics:8428

because it runs inside the container), and leave other settings for later.

After adding the source, you can query data on the instant query page:

Docker compose also launches a

categraf

collector container, allowing you to see metrics from one or more machines.

Import a dashboard: select a business group (e.g., Default Busi Group) and click the Import button.

Choose the Linux category and import all related dashboards; some may not work with the

categraf

collector, but you can test each to find suitable ones.

The "Linux Host by Categraf v2" dashboard shows all machines in the time‑series database. Dashboards can be filtered to show only machines belonging to the current business group.

Import alert rules: go to the alert rules page, click Import , select the Linux category, apply to all data sources, and enable later.

Configure a DingTalk bot: use keyword authentication (keyword "20" works because alerts contain timestamps). Batch‑update the imported alert rules with the DingTalk webhook URL.

After enabling the rules, alerts will be sent to DingTalk; you can also manually lower a rule's threshold to trigger an alert quickly.

For more features, refer to the official Nightingale documentation.