Hi ClickHouse community,

I'm the maintainer of Glances — a cross-platform,
open-source system monitoring tool written in Python (32k+ GitHub stars). It collects real-time
metrics from CPU, memory, disk, network, processes, containers, and more, and exposes them via
a terminal UI, a web interface, and a REST API.

One of Glances' core features is its export layer: collected metrics can be pushed
continuously to external backends. Existing targets include InfluxDB, Prometheus, Elasticsearch,
Graphite, Cassandra, MongoDB, Kafka, MQTT, and others.

Starting with the upcoming v4.5.3 (currently on the develop branch), ClickHouse joins that
list.

How it works

Glances uses the clickhouse-connect Python
client. Configuration is minimal — add a [clickhouse] section to glances.conf:

[clickhouse]
host=localhost
port=8123
db=glances
user=default
password=password

Then start Glances with:

glances --export clickhouse

All active plugin metrics are written to ClickHouse in real time. Given ClickHouse's strength in
time-series ingestion and analytical queries, it's a natural fit for long-term system monitoring
at scale — something a lot of Glances users who operate infrastructure for larger organisations
have been asking for.

Documentation

→ https://glances.readthedocs.io (will be updated on release)
→ Source doc (develop branch): https://github.com/nicolargo/glances/blob/develop/docs/gw/clickhouse.rst

Feedback welcome

If you're already using ClickHouse for infrastructure metrics and want to try this out before
the stable release, the develop branch is installable via:

pip install git+https://github.com/nicolargo/glances.git@develop

Happy to answer questions here or over on the Glances GitHub Discussions.