Package: prometheus-blackbox-exporter Version: 0.26.0-1 Severity: wishlist Hello Daniel,
I was also looking into setting up a blackbox exporter and ran into yet another issue related to logging. In its default configuration, the blackbox exporter logs messages at level INFO or higher. Let me give an example: time=2026-02-04T08:00:25.004Z level=INFO source=handler.go:122 msg="Beginning probe" module=icmp target=REDACTED probe=icmp timeout_seconds=5 time=2026-02-04T08:00:25.004Z level=INFO source=utils.go:61 msg="Resolving target address" module=icmp target=REDACTED target=REDACTED ip_protocol=ip6 time=2026-02-04T08:00:25.004Z level=INFO source=utils.go:107 msg="Resolved target address" module=icmp target=REDACTED target=REDACTED ip=REDACTED time=2026-02-04T08:00:25.004Z level=INFO source=icmp.go:108 msg="Creating socket" module=icmp target=REDACTED time=2026-02-04T08:00:25.005Z level=INFO source=icmp.go:218 msg="Creating ICMP packet" module=icmp target=REDACTED seq=19010 id=20069 time=2026-02-04T08:00:25.005Z level=INFO source=icmp.go:232 msg="Writing out packet" module=icmp target=REDACTED time=2026-02-04T08:00:25.005Z level=INFO source=icmp.go:306 msg="Waiting for reply packets" module=icmp target=REDACTED time=2026-02-04T08:00:25.197Z level=INFO source=icmp.go:371 msg="Found matching reply packet" module=icmp target=REDACTED time=2026-02-04T08:00:25.197Z level=INFO source=handler.go:133 msg="Probe succeeded" module=icmp target=REDACTED duration_seconds=0.192548366 I suggest that the verbosity of these logs is high. Personally, I would classify most of these entries as debug output. Running the exporter at this level with a couple of machines and services being promped results in 2GB worth of systemd-journal being filled in three hours. That's a lot. How about adding --log.level=warn via the default file installed by the package? If you disagree with this proposal, please tag the bug wontfix and close it. There is no point in debating. The setting is easily adapted. Helmut
