Hello,
I am requesting auto-connection for three interfaces across the valkey, valkey-chiseled and valkey-charmed snaps. These are required for proper initialization, performance health checks, and metrics exporting.
-
hardware-observe- Reason: Required by the
valkey-serverbinary to read/sys/devices/system/clocksource/.../current_clocksource. As an extreme-low-latency data store, Valkey checks this during startup to warn administrators if the system is using a slow virtualized clock instead of a fast hardware counter, which severely degrades throughput. - Log:
apparmor="DENIED" operation="open" class="file" profile="snap.valkey-chiseled.server" name="/sys/devices/system/clocksource/clocksource0/current_clocksource" pid=2983713 comm="valkey-server" requested_mask="r" denied_mask="r"
- Reason: Required by the
-
mount-observe- Reason: Required by
snap-exec(the Snap sandbox execution wrapper), as well as thevalkey-cliand other bundled tools. Whenever a snap binary, CLI tool, or hook is launched, the environment reads/proc/[pid]/mountinfoto map out the current mount namespace and verify the sandbox filesystem topology. - Log:
apparmor="DENIED" operation="open" class="file" profile="snap.valkey-chiseled.cli" name="/proc/2940903/mountinfo" pid=2940903 comm="snap-exec" requested_mask="r" denied_mask="r"
- Reason: Required by
-
network-observe- Reason: Required by the integrated Prometheus
metrics-exporter(Prometheus Redis exporter) invalkey-charmedto read/proc/[pid]/net/netstat. This is standard exporter behavior required to successfully scrape and expose network interface telemetry. - Log:
apparmor="DENIED" operation="open" class="file" profile="snap.valkey-charmed.metrics-exporter" name="/proc/3229994/net/netstat" pid=3229994 comm="prometheus-redi" requested_mask="r" denied_mask="r"
- Reason: Required by the integrated Prometheus
Please let me know if you need any additional details. Thank you!