We just released Onionprobe 1.3.0, a tool for testing and
monitoring the status of Onion Services.
This release fixes a security issue. Please upgrade as soon as possible!
ChangeLog
Fixes
- Standalone monitoring node:
-
SECURITY: use an internal network by default:
- By default, Prometheus, Alertmanager, Grafana and the Onionprobe exporter are accessible
only from localhost or through Onion Services. - This prevents exposing these services to the internet in systems that aren’t firewalled.
- This can be customized via the
ONIONPROBE_LISTEN_ADDR
environment variable used by
the Compose configuration. - Thanks to @gus for spotting the issue.
- By default, Prometheus, Alertmanager, Grafana and the Onionprobe exporter are accessible
-
The
start
action in theonionprobe-monitor
script now pulls and builds
images. -
Failure rate was erroneously being reported at 1% when all services
were working, when the excepted would be a reported value of 0%.
This is now fixed.
-
Features
-
- Grafana dashboard got a new time series: number of missing Onion Service
descriptors in HSDirs.
- Grafana dashboard got a new time series: number of missing Onion Service
-
Added tests for the Prometheus configuration
(tpo/onion-services/onionprobe!90).