Fold alerts under their host, recurring maintenance, run playbooks from alerts, and an Updates page

- Servers can say what they run on (or it is worked out from the Proxmox node); when the host is
  down the guests' offline alerts fold under it: held back, not counted, not notified, still recorded.
- Recurring maintenance schedules made into ordinary windows a week ahead (maintenance_schedules).
- Run a Semaphore template limited to a server from its page or an alert, and watch the run.
- Agents report pending updates (apt, dnf, zypper, Windows Update); an Updates page lists them.
- Migrations 0024 and 0025.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
This commit is contained in:
bobbanandClaude Sonnet 5.5 committed 2026-10-17 01:35:15 +02:00
1 parent 44e393f172
commit 171511c6db
45 files changed
+6772 -44

No files matched your search

+47 -1
View File
@@ -182,6 +182,46 @@ collect_systemd_timers() {
done < <(jq -c '.[]' <<<"$timers_json")
}
# Updates waiting to be installed, as JSON in UPDATES_JSON ("null" when this host's package manager isn't one we know, or couldn't answer).
# Nothing is installed or refreshed: this reads what the package manager already knows about, so it is quick and changes nothing.
collect_updates() {
UPDATES_JSON="null"
local manager="" count="" security="null" reboot="false" out="" status=0
if command -v apt-get >/dev/null 2>&1; then
manager="apt"
# -s simulates: "Inst" lines are what an upgrade would install. Security ones come from a "-security" or "Security" archive.
if out=$(LC_ALL=C timeout 60 apt-get -s -o Debug::NoLocking=true upgrade 2>/dev/null); then
count=$(grep -c '^Inst ' <<<"$out" || true)
security=$(grep '^Inst ' <<<"$out" | grep -ci 'security' || true)
fi
[[ -f /var/run/reboot-required ]] && reboot="true"
elif command -v dnf >/dev/null 2>&1; then
manager="dnf"
# check-update exits 100 when there are updates, 0 when none, 1 on an error. --cacheonly: never goes to the network from here.
out=$(LC_ALL=C timeout 120 dnf -q check-update --cacheonly 2>/dev/null) || status=$?
if [[ "$status" -eq 0 || "$status" -eq 100 ]]; then
count=$(awk 'NF >= 3 && $1 ~ /^[A-Za-z0-9_.+-]+\.[A-Za-z0-9_]+$/ && /Obsoleting/ == 0 { n++ } /^Obsoleting/ { exit } END { print n + 0 }' <<<"$out")
security=$(LC_ALL=C timeout 120 dnf -q updateinfo list --security --cacheonly 2>/dev/null | grep -c '/Sec\.' || true)
fi
if command -v needs-restarting >/dev/null 2>&1; then
needs-restarting -r >/dev/null 2>&1 || reboot="true"
fi
elif command -v zypper >/dev/null 2>&1; then
manager="zypper"
if out=$(LC_ALL=C timeout 120 zypper -q --no-refresh list-updates 2>/dev/null); then
count=$(grep -c '^v |' <<<"$out" || true)
fi
[[ -f /run/reboot-needed ]] && reboot="true"
fi
if [[ -n "$manager" && "$count" =~ ^[0-9]+$ ]]; then
[[ "$security" =~ ^[0-9]+$ ]] || security="null"
UPDATES_JSON=$(jq -n --arg manager "$manager" --argjson count "$count" --argjson security "$security" --argjson reboot "$reboot" \
'{manager: $manager, count: $count, security: $security, reboot_required: $reboot}')
fi
}
collect_system_info() {
local ip_json="[]"
if command -v ip >/dev/null 2>&1; then
@@ -243,7 +283,12 @@ collect_system_info() {
')
fi
# Best-effort like the rest: a package manager that misbehaves must not cost the report.
UPDATES_JSON="null"
collect_updates || UPDATES_JSON="null"
SYSTEM_JSON=$(jq -n \
--argjson updates "$UPDATES_JSON" \
--argjson ip_addresses "$ip_json" \
--argjson listening_ports "$ports_json" \
--arg cpu_model "$cpu_model" \
@@ -257,7 +302,8 @@ collect_system_info() {
cpu: { model: $cpu_model, cores: $cpu_cores, load_percent: $cpu_load_percent },
memory: { total_bytes: $mem_total_bytes, used_bytes: $mem_used_bytes },
disks: $disks,
listening_ports: $listening_ports
listening_ports: $listening_ports,
updates: $updates
}')
}