Merge remote-tracking branch 'origin/3/dev' into saltthangs

Merge pull request #16033 from Security-Onion-Solutions/jertel/wip
avoid setup failure reason ambiguity
2026-07-03 07:38:16 +02:00 · 2026-07-02 10:27:54 -04:00 · 2026-07-02 09:20:48 -04:00 · 2026-07-02 09:04:57 -04:00 · 2026-07-02 08:59:52 -04:00 · 2026-07-01 16:54:02 -04:00
76 changed files with 1678 additions and 241 deletions
@@ -1,59 +0,0 @@
-#!/usr/bin/env bash
-
-# This script adds sensors/nodes/etc to the nodes tab
-default_salt_dir=/opt/so/saltstack/default
-local_salt_dir=/opt/so/saltstack/local
-TYPE=$1
-NAME=$2
-IPADDRESS=$3
-CPUS=$4
-GUID=$5
-MANINT=$6
-ROOTFS=$7
-NSM=$8
-MONINT=$9
-#NODETYPE=$10
-#HOTNAME=$11
-
-echo "Seeing if this host is already in here. If so delete it"
-if grep -q $NAME "$local_salt_dir/pillar/data/$TYPE.sls"; then
-  echo "Node Already Present - Let's re-add it"
-  awk -v blah="  $NAME:" 'BEGIN{ print_flag=1 }
-{
-    if( $0 ~ blah )
-    {
-       print_flag=0;
-       next
-    }
-    if( $0 ~ /^  [a-zA-Z0-9]+:$/ )
-    {
-        print_flag=1;
-    }
-    if ( print_flag == 1 )
-        print $0
-
-} ' $local_salt_dir/pillar/data/$TYPE.sls > $local_salt_dir/pillar/data/tmp.$TYPE.sls
-mv $local_salt_dir/pillar/data/tmp.$TYPE.sls $local_salt_dir/pillar/data/$TYPE.sls
-echo "Deleted $NAME from the tab. Now adding it in again with updated info"
-fi
-echo "  $NAME:" >> $local_salt_dir/pillar/data/$TYPE.sls
-echo "    ip: $IPADDRESS" >> $local_salt_dir/pillar/data/$TYPE.sls
-echo "    manint: $MANINT" >> $local_salt_dir/pillar/data/$TYPE.sls
-echo "    totalcpus: $CPUS" >> $local_salt_dir/pillar/data/$TYPE.sls
-echo "    guid: $GUID" >> $local_salt_dir/pillar/data/$TYPE.sls
-echo "    rootfs: $ROOTFS" >> $local_salt_dir/pillar/data/$TYPE.sls
-echo "    nsmfs: $NSM" >> $local_salt_dir/pillar/data/$TYPE.sls
-if [ $TYPE == 'sensorstab' ]; then
-  echo "    monint: bond0" >> $local_salt_dir/pillar/data/$TYPE.sls
-fi
-if [ $TYPE == 'evaltab' ] || [ $TYPE == 'standalonetab' ]; then
-  echo "    monint: bond0" >> $local_salt_dir/pillar/data/$TYPE.sls
-  if [ ! $10 ]; then
-    salt-call state.apply utility queue=True
-  fi
-fi
-if [ $TYPE == 'nodestab' ]; then
-  salt-call state.apply elasticsearch queue=True
-#  echo "    nodetype: $NODETYPE" >> $local_salt_dir/pillar/data/$TYPE.sls
-#  echo "    hotname: $HOTNAME" >> $local_salt_dir/pillar/data/$TYPE.sls
-fi
@@ -3,6 +3,8 @@ base:
    - ca
    - global.soc_global
    - global.adv_global
+    - salt.soc_salt
+    - salt.adv_salt
    - docker.soc_docker
    - docker.adv_docker
    - influxdb.token
@@ -0,0 +1,142 @@
+# Copyright Security Onion Solutions LLC and/or licensed to Security Onion Solutions LLC under one
+# or more contributor license agreements. Licensed under the Elastic License 2.0 as shown at
+# https://securityonion.net/license; you may not use this file except in compliance with the
+# Elastic License 2.0.
+
+# Custom salt beacon that watches the SOC audit_settings table in postgres for
+# new settings changes and emits a beacon event per new row. This replaces the
+# inotify watch on /opt/so/saltstack/local/pillar -- instead of monitoring pillar
+# files on disk, we monitor the securityonion.audit_settings table that SOC writes to.
+#
+# Detection is poll-based with a monotonic `id` watermark persisted to
+# WATERMARK_FILE: each pass selects rows with id greater than the last id seen,
+# which makes it self-healing (a missed poll simply catches up on the next one).
+#
+# Each emitted event carries setting_id and node_id; the push_pillar reactor maps
+# setting_id -> app via pillar_push_map.yaml and writes a push intent, after which
+# the existing so-push-drainer / orch.push_batch pipeline takes over unchanged.
+
+import logging
+import os
+import subprocess
+
+log = logging.getLogger(__name__)
+
+WATERMARK_FILE = '/opt/so/state/postgres_pillar_beacon_watch.id'
+CONTAINER = 'so-postgres'
+DATABASE = 'securityonion'
+
+# Unaligned, tuples-only psql output with a field separator that cannot appear in
+# an id/setting_id/node_id, so we can split each row reliably.
+FIELD_SEP = '\x1f'
+
+
+def __virtual__():
+    return True
+
+
+def validate(config):
+    return True, 'valid'
+
+
+def _read_watermark():
+    # Returns the last processed id, or None if the watermark has not been seeded.
+    try:
+        with open(WATERMARK_FILE, 'r') as f:
+            return int((f.read() or '').strip())
+    except (IOError, ValueError):
+        return None
+
+
+def _write_watermark(value):
+    try:
+        os.makedirs(os.path.dirname(WATERMARK_FILE), exist_ok=True)
+        tmp = WATERMARK_FILE + '.tmp'
+        with open(tmp, 'w') as f:
+            f.write(str(int(value)))
+        os.rename(tmp, WATERMARK_FILE)
+    except OSError:
+        log.exception('postgres_pillar_beacon: failed to persist watermark to %s', WATERMARK_FILE)
+
+
+def _query(sql):
+    # Run a query against securityonion inside the so-postgres container over the unix
+    # socket (trust auth, no password). Returns stdout on success, or None on any
+    # failure so the caller can no-op and retry on the next interval.
+    cmd = [
+        'docker', 'exec', CONTAINER,
+        'psql', '-U', 'postgres', '-d', DATABASE,
+        '-tA', '-F', FIELD_SEP, '-c', sql,
+    ]
+    try:
+        result = subprocess.run(cmd, capture_output=True, text=True, timeout=30)
+    except subprocess.TimeoutExpired:
+        log.warning('postgres_pillar_beacon: psql timed out')
+        return None
+    except Exception:
+        log.exception('postgres_pillar_beacon: failed to exec psql')
+        return None
+    if result.returncode != 0:
+        log.warning('postgres_pillar_beacon: psql failed (rc=%s): %s',
+                    result.returncode, (result.stderr or '').strip())
+        return None
+    return result.stdout
+
+
+def beacon(config):
+    retval = []
+
+    watermark = _read_watermark()
+
+    # First run / missing watermark: seed to the current MAX(id) and emit nothing
+    # so we never replay the entire settings history into a fleetwide push.
+    if watermark is None:
+        seed = _query('SELECT COALESCE(MAX(id), 0) FROM audit_settings;')
+        if seed is None:
+            return retval  # postgres not ready yet; retry next interval
+        try:
+            _write_watermark(int((seed or '0').strip() or 0))
+        except ValueError:
+            log.warning('postgres_pillar_beacon: could not parse MAX(id) seed: %r', seed)
+        return retval
+
+    rows = _query(
+        "SELECT id, setting_id, COALESCE(node_id, '') FROM audit_settings "
+        "WHERE id > %d ORDER BY id;" % watermark
+    )
+    if rows is None:
+        return retval
+
+    max_id = watermark
+    for line in rows.splitlines():
+        # Do NOT str.strip() the whole line: Python treats the \x1f field
+        # separator (and \x1c-\x1e) as whitespace, so stripping would eat an
+        # empty trailing node_id field and make the row look malformed.
+        if not line.strip():
+            continue
+        parts = line.split(FIELD_SEP)
+        if len(parts) < 3:
+            log.warning('postgres_pillar_beacon: skipping malformed row: %r', line)
+            continue
+        try:
+            row_id = int(parts[0])
+        except ValueError:
+            log.warning('postgres_pillar_beacon: skipping row with non-int id: %r', line)
+            continue
+        setting_id = parts[1]
+        node_id = parts[2]
+        retval.append({
+            'tag': 'audit_settings',
+            'id': row_id,
+            'setting_id': setting_id,
+            'node_id': node_id,
+        })
+        if row_id > max_id:
+            max_id = row_id
+
+    if max_id > watermark:
+        _write_watermark(max_id)
+        log.info('postgres_pillar_beacon: emitted %d change(s), watermark %d -> %d',
+                 len(retval), watermark, max_id)
+
+    return retval
@@ -0,0 +1,139 @@
+# Copyright Security Onion Solutions LLC and/or licensed to Security Onion Solutions LLC under one
+# or more contributor license agreements. Licensed under the Elastic License 2.0 as shown at
+# https://securityonion.net/license; you may not use this file except in compliance with the
+# Elastic License 2.0.
+
+# Custom salt beacon that watches the suricata/strelka rule directories for changes
+# and emits a beacon event per changed directory. This replaces the stock salt
+# `inotify` beacon, which leaks a kernel inotify instance every time the minion
+# rebuilds the beacon loader's __context__ (orphaning the old pyinotify.Notifier
+# without closing it) until fs.inotify.max_user_instances is exhausted and the
+# beacon dies with EMFILE. Polling holds zero inotify instances, so the leak is
+# impossible, and it keeps firing during state runs (no blackout).
+#
+# Detection is poll-based with a per-directory fingerprint persisted to
+# WATERMARK_DIR: each pass walks the directory and hashes every file's
+# (relpath, st_mtime_ns, st_size), which catches content writes, additions,
+# moves, and deletions. A change in the digest emits one event; an unchanged
+# digest emits nothing. This makes it self-healing (a missed poll simply catches
+# up on the next one).
+#
+# Each emitted event carries the watched directory path under the configured tag
+# (e.g. salt/beacon/<minion>/rules_beacon/suricata); the push_suricata / push_strelka
+# reactors write a push intent, after which the existing so-push-drainer /
+# orch.push_batch pipeline takes over unchanged.
+
+import hashlib
+import logging
+import os
+import re
+
+log = logging.getLogger(__name__)
+
+WATERMARK_DIR = '/opt/so/state'
+
+# Temp/editor files that should not trigger a push. Mirrors the exclude regexes
+# the inotify beacon used. Matched against the full pathname.
+EXCLUDES = [
+    re.compile(r'\.sw[a-z]$'),
+    re.compile(r'~$'),
+    re.compile(r'/4913$'),
+    re.compile(r'/\.#'),
+]
+
+
+def __virtual__():
+    return True
+
+
+def validate(config):
+    return True, 'valid'
+
+
+def _paths_from_config(config):
+    # The beacon config arrives as a list of single-key dicts (salt beacon style).
+    # Merge it and return the {dir: tag} mapping under the 'paths' key.
+    merged = {}
+    if isinstance(config, list):
+        for item in config:
+            if isinstance(item, dict):
+                merged.update(item)
+    elif isinstance(config, dict):
+        merged = config
+    paths = merged.get('paths', {})
+    return paths if isinstance(paths, dict) else {}
+
+
+def _excluded(pathname):
+    for pattern in EXCLUDES:
+        if pattern.search(pathname):
+            return True
+    return False
+
+
+def _fingerprint(directory):
+    # Stat-only walk; hash each file's (relpath, mtime_ns, size). Returns a hex
+    # digest, or the digest of an empty tree if the directory does not exist.
+    h = hashlib.sha1()
+    if os.path.isdir(directory):
+        entries = []
+        for root, _dirs, files in os.walk(directory):
+            for name in files:
+                full = os.path.join(root, name)
+                if _excluded(full):
+                    continue
+                try:
+                    st = os.stat(full)
+                except OSError:
+                    continue
+                rel = os.path.relpath(full, directory)
+                entries.append('%s\0%d\0%d' % (rel, st.st_mtime_ns, st.st_size))
+        for line in sorted(entries):
+            h.update(line.encode('utf-8', 'surrogateescape'))
+            h.update(b'\n')
+    return h.hexdigest()
+
+
+def _watermark_file(tag):
+    return os.path.join(WATERMARK_DIR, 'rules_beacon_%s.hash' % tag)
+
+
+def _read_watermark(tag):
+    try:
+        with open(_watermark_file(tag), 'r') as f:
+            return (f.read() or '').strip() or None
+    except IOError:
+        return None
+
+
+def _write_watermark(tag, digest):
+    path = _watermark_file(tag)
+    try:
+        os.makedirs(WATERMARK_DIR, exist_ok=True)
+        tmp = path + '.tmp'
+        with open(tmp, 'w') as f:
+            f.write(digest)
+        os.rename(tmp, path)
+    except OSError:
+        log.exception('rules_beacon: failed to persist watermark to %s', path)
+
+
+def beacon(config):
+    retval = []
+
+    for directory, tag in _paths_from_config(config).items():
+        digest = _fingerprint(directory)
+        previous = _read_watermark(tag)
+
+        # First run / missing watermark: seed the digest and emit nothing so a
+        # fresh host does not fire a spurious fleetwide push.
+        if previous is None:
+            _write_watermark(tag, digest)
+            continue
+
+        if digest != previous:
+            _write_watermark(tag, digest)
+            retval.append({'tag': tag, 'path': directory})
+            log.info('rules_beacon: change detected in %s, emitting %s', directory, tag)
+
+    return retval
@@ -37,8 +37,7 @@
    'elasticfleet',
    'elasticfleet.manager',
    'elasticsearch.cluster',
-    'elastic-fleet-package-registry',
-    'utility'
+    'elastic-fleet-package-registry'
 ] %}

 {% set sensor_states = [
@@ -291,6 +291,20 @@ download_and_verify() {
 	fi
 }

+# check if container with name is running and optionally stop it
+docker_check_running() {
+    # show running containers, only names
+    if docker ps --format '{{.Names}}' | grep -q "^so-${1}$"; then
+        if [[ "$2" == "--stop" ]]; then
+            docker stop "so-${1}"
+        fi
+
+        return 0
+    else
+        return 1
+    fi
+}
+
 elastic_license() {

 read -r -d '' message <<- EOM
@@ -5,27 +5,41 @@
 # https://securityonion.net/license; you may not use this file except in compliance with the
 # Elastic License 2.0.

-
-
-# Usage: so-restart  kibana | playbook 
-
 . /usr/sbin/so-common

-if [ $# -ge 1 ]; then
+usage() {
+    echo "Usage: $0 <component> [args]"
+    echo ""
+    echo "Supported args:"
+    echo "  --force | -f          Force stop all Salt jobs before starting component."
+    echo ""
+    echo "Examples:"
+    echo "  $0 kibana             Restart Kibana"
+    echo "  $0 kibana --force     Force stop all Salt jobs before restarting Kibana"
+    exit 1
+}

-        echo $banner
-        printf "Restarting $1...\n\nThis could take a while if another Salt job is running. \nRun this command with --force to stop all Salt jobs before proceeding.\n"
-        echo $banner
-
-        if [ "$2" = "--force" ]; then
-                printf "\nForce-stopping all Salt jobs before proceeding\n\n"
-                salt-call saltutil.kill_all_jobs
-        fi
-
-        case $1 in
-                "elastic-fleet") docker stop so-elastic-fleet && docker rm so-elastic-fleet && salt-call state.apply elasticfleet queue=True;;
-                *)  docker stop so-$1 ; docker rm so-$1 ; salt-call state.apply $1 queue=True;;
-        esac
-else
-        echo -e "\nPlease provide an argument by running like so-restart $component, or by using the component-specific script.\nEx. so-restart logstash, or so-logstash-restart\n"
+if [[ $# -lt 1 ]]; then
+    usage
 fi
+
+#shellcheck disable=SC2154
+echo "$banner"
+printf "Restarting %s...\n\nThis could take a while if another Salt job is running. \nRun this command with --force to stop all Salt jobs before proceeding.\n" "$1"
+echo "$banner"
+if [[ "$2" = "--force" ]] || [[ "$2" = "-f" ]]; then
+    printf "\nForce-stopping all Salt jobs before proceeding\n\n"
+    salt-call saltutil.kill_all_jobs
+fi
+case $1 in
+    "elastic-fleet"|"elasticfleet")
+        docker_check_running "elastic-fleet" "--stop"
+        docker rm "so-elastic-fleet" 2> /dev/null
+        salt-call state.apply elasticfleet queue=True
+        ;;
+    *)
+        docker_check_running "$1" "--stop"
+        docker rm "so-${1}" 2> /dev/null
+        salt-call state.apply "$1" queue=True
+        ;;
+esac
@@ -5,27 +5,54 @@
 # https://securityonion.net/license; you may not use this file except in compliance with the
 # Elastic License 2.0.

-
-
-# Usage: so-start  all | kibana | playbook 
-
+# shellcheck disable=SC1091
 . /usr/sbin/so-common

-if [ $# -ge 1 ]; then
-	echo $banner
-	printf "Starting $1...\n\nThis could take a while if another Salt job is running. \nRun this command with --force to stop all Salt jobs before proceeding.\n"
-	echo $banner
+usage() {
+    echo "Usage: $0 <component> [args]"
+    echo ""
+    echo "Supported args:"
+    echo "  --force | -f          Force stop all Salt jobs before starting component."
+    echo ""
+    echo "Examples:"
+    echo "  $0 kibana             Start Kibana"
+    echo "  $0 kibana --force     Force stop all Salt jobs before starting Kibana"
+    exit 1
+}

-	if [ "$2" = "--force" ]; then
-   		printf "\nForce-stopping all Salt jobs before proceeding\n\n"
-   		salt-call saltutil.kill_all_jobs
-	fi
-
-	case $1 in
-   		"all") salt-call state.highstate queue=True;;
-   		"elastic-fleet") if docker ps | grep -q so-$1; then printf "\n$1 is already running!\n\n"; else docker rm so-$1 >/dev/null 2>&1 ; salt-call state.apply elasticfleet queue=True; fi ;; 
-   		*) if docker ps | grep -E -q '^so-$1$'; then printf "\n$1 is already running\n\n"; else docker rm so-$1 >/dev/null 2>&1 ; salt-call state.apply $1 queue=True; fi ;; 
-	esac
-else
-	echo -e "\nPlease provide an argument by running like so-start $component, or by using the component-specific script.\nEx. so-start logstash, or so-logstash-start\n"	
+if [[ $# -lt 1 ]]; then
+    usage
 fi
+
+#shellcheck disable=SC2154
+echo "$banner"
+printf "Starting %s...\n\nThis could take a while if another Salt job is running. \nRun this command with --force to stop all Salt jobs before proceeding.\n" "$1"
+echo "$banner"
+if [[ "$2" = "--force" ]] || [[ "$2" == "-f" ]]; then
+	printf "\nForce-stopping all Salt jobs before proceeding\n\n"
+	salt-call saltutil.kill_all_jobs
+fi
+
+case "$1" in
+    "all")
+        salt-call state.highstate queue=True
+        ;;
+   	"elastic-fleet"|"elasticfleet")
+        if docker_check_running "elastic-fleet"; then
+            printf "\nso-%s is already running!\n\n" "elastic-fleet"
+            /usr/sbin/so-status
+        else
+            docker rm "so-elastic-fleet" 2> /dev/null
+            salt-call state.apply elasticfleet queue=True
+        fi
+        ;;
+   	*)
+        if docker_check_running "$1"; then
+            printf "\nso-%s is already running\n\n" "$1"
+            /usr/sbin/so-status
+        else
+            docker rm "so-${1}" 2> /dev/null
+            salt-call state.apply "$1" queue=True
+        fi
+        ;;
+esac
@@ -5,21 +5,33 @@
 # https://securityonion.net/license; you may not use this file except in compliance with the
 # Elastic License 2.0.

-
-
-# Usage: so-stop kibana | playbook | thehive
-
+# shellcheck disable=SC1091
 . /usr/sbin/so-common

-if [ $# -ge 1 ]; then
-	echo $banner
-	printf "Stopping $1...\n"
-	echo $banner
+usage() {
+    echo "Usage: $0 <component>"
+    echo ""
+    echo "Examples:"
+    echo "  $0 kibana             Stop Kibana"
+    exit 1
+}

-	case $1 in
-    		*) docker stop so-$1 ; docker rm so-$1 ;;
-	esac
-else
-	echo -e "\nPlease provide an argument by running like so-stop $component, or by using the component-specific script.\nEx. so-stop logstash, or so-logstash-stop\n"	
+if [[ $# -lt 1 ]]; then
+    usage
 fi

+
+#shellcheck disable=SC2154
+echo "$banner"
+printf "Stopping %s...\n" "$1"
+echo "$banner"
+case $1 in
+	"elasticfleet"|"elastic-fleet")
+        docker_check_running "elastic-fleet" "--stop"
+        docker rm "so-elastic-fleet" 2> /dev/null
+        ;;
+    *)
+        docker_check_running "$1" "--stop"
+        docker rm "so-${1}" 2> /dev/null
+    ;;
+esac
@@ -1,5 +1,3 @@
-{% import_yaml 'salt/minion.defaults.yaml' as SALT_MINION_DEFAULTS -%}
-
 #!/bin/bash
 #
 # Copyright Security Onion Solutions LLC and/or licensed to Security Onion Solutions LLC under one
@@ -7,7 +5,7 @@
 # https://securityonion.net/license; you may not use this file except in compliance with the
 # Elastic License 2.0.

-
+{% from 'salt/schedule.map.jinja' import SCHEDULEMERGED %}

 # this script checks the time the file /opt/so/log/salt/state-apply-test was last modified and restarts the salt-minion service if it is outside a threshold date/time
 # the file is modified via file.touch using a scheduled job healthcheck.salt-minion.state-apply-test that runs a state.apply. 
@@ -25,7 +23,8 @@ SYSTEM_START_TIME=$(date -d "$(</proc/uptime awk '{print $1}') seconds ago" +%s)
 LAST_HIGHSTATE_END=$([ -e "/opt/so/log/salt/lasthighstate" ] && date -r /opt/so/log/salt/lasthighstate +%s || echo 0)
 LAST_HEALTHCHECK_STATE_APPLY=$([ -e "/opt/so/log/salt/state-apply-test" ] && date -r /opt/so/log/salt/state-apply-test +%s || echo 0)
 # SETTING THRESHOLD TO ANYTHING UNDER 600 seconds may cause a lot of salt-minion restarts since the job to touch the file occurs every 5-8 minutes by default
-THRESHOLD={{SALT_MINION_DEFAULTS.salt.minion.check_threshold}} #within how many seconds the file /opt/so/log/salt/state-apply-test must have been touched/modified before the salt minion is restarted
+# THRESHOLD is derived from the salt schedule highstate interval + 1 hour, so the minion-check grace period tracks the schedule automatically.
+THRESHOLD=$(( ({{ SCHEDULEMERGED.highstate_interval_hours }} + 1) * 3600 )) #within how many seconds the file /opt/so/log/salt/state-apply-test must have been touched/modified before the salt minion is restarted
 THRESHOLD_DATE=$((LAST_HEALTHCHECK_STATE_APPLY+THRESHOLD))

 logCmd() {
@@ -9,7 +9,8 @@
 prune_images:
  cmd.run:
    - name: so-docker-prune
-    - order: last
+    - onlyif: command -v /usr/sbin/so-docker-prune >/dev/null 2>&1
+    - order: 9000

 {% else %}

@@ -19,6 +19,7 @@ wait_for_elasticsearch:
 so-elastalert:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-elastalert:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - hostname: elastalert
    - name: so-elastalert
    - user: so-elastalert
@@ -15,6 +15,7 @@ include:
 so-elastic-fleet-package-registry:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-elastic-fleet-package-registry:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - name: so-elastic-fleet-package-registry
    - hostname: Fleet-package-reg-{{ GLOBALS.hostname }}
    - detach: True
@@ -16,6 +16,7 @@ include:
 so-elastic-agent:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-elastic-agent:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - name: so-elastic-agent
    - hostname: {{ GLOBALS.hostname }}
    - detach: True
@@ -42,6 +42,7 @@ elasticagent_syncartifacts:
 so-elastic-fleet:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-elastic-agent:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - name: so-elastic-fleet
    - hostname: FleetServer-{{ GLOBALS.hostname }}
    - detach: True
@@ -24,6 +24,7 @@ include:
 so-elasticsearch:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-elasticsearch:{{ ELASTICSEARCHMERGED.version }}
+    - restart_policy: unless-stopped
    - hostname: elasticsearch
    - name: so-elasticsearch
    - user: elasticsearch
@@ -1,3 +1,3 @@
 global:
  pcapengine: SURICATA
-  pipeline: REDIS
+  pipeline: REDIS
@@ -58,6 +58,7 @@ so-hydra:
      - {{ ULIMIT.name }}={{ ULIMIT.soft }}:{{ ULIMIT.hard }}
    {%   endfor %}
    {% endif %}
+    # Intentionally unless-stopped -- matches the fleet default.
    - restart_policy: unless-stopped
    - watch:
      - file: hydraconfig
@@ -15,6 +15,7 @@ include:
 so-idh:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-idh:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - name: so-idh
    - detach: True
    - network_mode: host
@@ -18,6 +18,7 @@ include:
 so-influxdb:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-influxdb:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - hostname: influxdb
    - networks:
      - sobridge:
@@ -27,6 +27,7 @@ include:
 so-kafka:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-kafka:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - hostname: so-kafka
    - name: so-kafka
    - networks:
@@ -17,6 +17,7 @@ include:
 so-kibana:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-kibana:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - hostname: kibana
    - user: "932:0"
    - networks:
@@ -69,7 +70,7 @@ wait_for_so-kibana:
    - ssl: True
    - verify_ssl: False
    - status: 200
-    - wait_for: 300
+    - wait_for: 600
    - request_interval: 15
    - require:
      - docker_container: so-kibana
@@ -51,6 +51,7 @@ so-kratos:
      - {{ ULIMIT.name }}={{ ULIMIT.soft }}:{{ ULIMIT.hard }}
    {%   endfor %}
    {% endif %}
+    # Intentionally unless-stopped -- matches the fleet default.
    - restart_policy: unless-stopped
    - watch:
      - file: kratosschema
@@ -28,6 +28,7 @@ include:
 so-logstash:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-logstash:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - hostname: so-logstash
    - name: so-logstash
    - networks:
@@ -0,0 +1,21 @@
+{% from 'vars/globals.map.jinja' import GLOBALS %}
+{% from 'salt/auto_apply.map.jinja' import AUTOAPPLY %}
+
+include:
+  - salt.minion
+
+{% if GLOBALS.is_manager and AUTOAPPLY.enabled %}
+salt_beacons_pushstate:
+  file.managed:
+    - name: /etc/salt/minion.d/beacons_pushstate.conf
+    - source: salt://manager/files/beacons_pushstate.conf.jinja
+    - template: jinja
+    - watch_in:
+      - service: salt_minion_service
+{% else %}
+salt_beacons_pushstate:
+  file.absent:
+    - name: /etc/salt/minion.d/beacons_pushstate.conf
+    - watch_in:
+      - service: salt_minion_service
+{% endif %}
@@ -0,0 +1,11 @@
+{% from 'salt/auto_apply.map.jinja' import AUTOAPPLY %}
+beacons:
+  postgres_pillar_beacon:
+    - interval: {{ AUTOAPPLY.drain_interval }}
+    - disable_during_state_run: False
+  rules_beacon:
+    - interval: {{ AUTOAPPLY.drain_interval }}
+    - disable_during_state_run: False
+    - paths:
+        /opt/so/saltstack/local/salt/suricata/rules: suricata
+        /opt/so/saltstack/local/salt/strelka/rules/compiled: strelka
@@ -11,8 +11,8 @@ name=Security Onion Repo repo
 mirrorlist=file:///opt/so/conf/reposync/mirror.txt
 enabled=1
 gpgcheck=1
-[securityonionkernel]
-name=Security Onion Repo repo
+[securityonionkernelsync]
+name=Security Onion Kernel Repo repo
 mirrorlist=file:///opt/so/conf/reposync/mirror-kernel.txt
 enabled=1
 gpgcheck=1
@@ -15,6 +15,7 @@ include:
  - manager.elasticsearch
  - manager.kibana
  - manager.managed_soc_annotations
+  - manager.beacons

 repo_log_dir:
  file.directory:
@@ -260,6 +261,7 @@ surifiltersrules:
    - user: 939
    - group: 939

+
 {% else %}

 {{sls}}_state_not_allowed:
@@ -0,0 +1,232 @@
+#!/opt/saltstack/salt/bin/python3
+
+# Copyright Security Onion Solutions LLC and/or licensed to Security Onion Solutions LLC under one
+# or more contributor license agreements. Licensed under the Elastic License 2.0 as shown at
+# https://securityonion.net/license; you may not use this file except in compliance with the
+# Elastic License 2.0.
+
+"""
+so-push-drainer
+===============
+
+Scheduled drainer for the active-push feature. Runs on the manager every
+drain_interval seconds (default 15) via a salt schedule in salt/salt/push_drain_schedule.sls.
+
+For each intent file under /opt/so/state/push_pending/*.json whose last_touch
+is older than debounce_seconds, this script:
+  * concatenates the actions lists from every ready intent
+  * dedupes by (state or __highstate__, tgt, tgt_type)
+  * dispatches a single `salt-run state.orchestrate orch.push_batch --async`
+    with the deduped actions list passed as pillar kwargs
+  * deletes the contributed intent files on successful dispatch
+
+Reactor sls files (push_suricata, push_strelka, push_pillar) write intents
+but never dispatch directly -- see plan
+/home/mreeves/.claude/plans/goofy-marinating-hummingbird.md for the full design.
+"""
+
+import fcntl
+import glob
+import json
+import logging
+import logging.handlers
+import os
+import subprocess
+import sys
+import time
+
+import salt.client
+
+PENDING_DIR = '/opt/so/state/push_pending'
+LOCK_FILE = os.path.join(PENDING_DIR, '.lock')
+LOG_FILE = '/opt/so/log/salt/so-push-drainer.log'
+
+HIGHSTATE_SENTINEL = '__highstate__'
+
+
+def _make_logger():
+    logger = logging.getLogger('so-push-drainer')
+    logger.setLevel(logging.INFO)
+    if not logger.handlers:
+        os.makedirs(os.path.dirname(LOG_FILE), exist_ok=True)
+        handler = logging.handlers.RotatingFileHandler(
+            LOG_FILE, maxBytes=5 * 1024 * 1024, backupCount=3,
+        )
+        handler.setFormatter(logging.Formatter(
+            '%(asctime)s | %(levelname)s | %(message)s',
+        ))
+        logger.addHandler(handler)
+    return logger
+
+
+def _load_push_cfg():
+    """Read the salt:auto_apply pillar subtree via salt-call. Returns a dict."""
+    caller = salt.client.Caller()
+    cfg = caller.cmd('pillar.get', 'salt:auto_apply', {})
+    return cfg if isinstance(cfg, dict) else {}
+
+
+def _read_intent(path, log):
+    try:
+        with open(path, 'r') as f:
+            return json.load(f)
+    except (IOError, ValueError) as exc:
+        log.warning('cannot read intent %s: %s', path, exc)
+        return None
+    except Exception:
+        log.exception('unexpected error reading %s', path)
+        return None
+
+
+def _dedupe_actions(actions):
+    seen = set()
+    deduped = []
+    for action in actions:
+        if not isinstance(action, dict):
+            continue
+        state_key = HIGHSTATE_SENTINEL if action.get('highstate') else action.get('state')
+        tgt = action.get('tgt')
+        tgt_type = action.get('tgt_type', 'compound')
+        if not state_key or not tgt:
+            continue
+        key = (state_key, tgt, tgt_type)
+        if key in seen:
+            continue
+        seen.add(key)
+        deduped.append(action)
+    return deduped
+
+
+def _dispatch(actions, log):
+    pillar_arg = json.dumps({'actions': actions})
+    cmd = [
+        'salt-run',
+        'state.orchestrate',
+        'orch.push_batch',
+        'pillar={}'.format(pillar_arg),
+        '--async',
+    ]
+    log.info('dispatching: %s', ' '.join(cmd[:3]) + ' pillar=<{} actions>'.format(len(actions)))
+    try:
+        result = subprocess.run(
+            cmd, check=True, capture_output=True, text=True, timeout=60,
+        )
+    except subprocess.CalledProcessError as exc:
+        log.error('dispatch failed (rc=%s): stdout=%s stderr=%s',
+                  exc.returncode, exc.stdout, exc.stderr)
+        return False
+    except subprocess.TimeoutExpired:
+        log.error('dispatch timed out after 60s')
+        return False
+    except Exception:
+        log.exception('dispatch raised')
+        return False
+    log.info('dispatch accepted: %s', (result.stdout or '').strip())
+    return True
+
+
+def main():
+    log = _make_logger()
+
+    if not os.path.isdir(PENDING_DIR):
+        # Nothing to do; reactors create the dir on first use.
+        return 0
+
+    try:
+        push = _load_push_cfg()
+    except Exception:
+        log.exception('failed to read salt:auto_apply pillar; aborting drain pass')
+        return 1
+
+    if not push.get('enabled', True):
+        log.debug('push disabled; exiting')
+        return 0
+
+    debounce_seconds = int(push.get('debounce_seconds', 30))
+
+    os.makedirs(PENDING_DIR, exist_ok=True)
+    lock_fd = os.open(LOCK_FILE, os.O_CREAT | os.O_RDWR, 0o644)
+    try:
+        fcntl.flock(lock_fd, fcntl.LOCK_EX)
+
+        intent_files = [
+            p for p in sorted(glob.glob(os.path.join(PENDING_DIR, '*.json')))
+            if os.path.basename(p) != '.lock'
+        ]
+        if not intent_files:
+            return 0
+
+        now = time.time()
+        ready = []
+        skipped = 0
+        broken = []
+        for path in intent_files:
+            intent = _read_intent(path, log)
+            if not isinstance(intent, dict):
+                broken.append(path)
+                continue
+            last_touch = intent.get('last_touch', 0)
+            if now - last_touch < debounce_seconds:
+                skipped += 1
+                continue
+            ready.append((path, intent))
+
+        for path in broken:
+            try:
+                os.unlink(path)
+            except OSError:
+                pass
+
+        if not ready:
+            if skipped:
+                log.debug('no ready intents (%d still in debounce window)', skipped)
+            return 0
+
+        combined_actions = []
+        oldest_first_touch = now
+        all_paths = []
+        for path, intent in ready:
+            combined_actions.extend(intent.get('actions', []) or [])
+            first = intent.get('first_touch', now)
+            if first < oldest_first_touch:
+                oldest_first_touch = first
+            all_paths.extend(intent.get('paths', []) or [])
+
+        deduped = _dedupe_actions(combined_actions)
+        if not deduped:
+            log.warning('%d intent(s) had no usable actions; clearing', len(ready))
+            for path, _ in ready:
+                try:
+                    os.unlink(path)
+                except OSError:
+                    pass
+            return 0
+
+        debounce_duration = now - oldest_first_touch
+        log.info(
+            'draining %d intent(s): %d action(s) after dedupe (raw=%d), '
+            'debounce_duration=%.1fs, paths=%s',
+            len(ready), len(deduped), len(combined_actions),
+            debounce_duration, all_paths[:20],
+        )
+
+        if not _dispatch(deduped, log):
+            log.warning('dispatch failed; leaving intent files in place for retry')
+            return 1
+
+        for path, _ in ready:
+            try:
+                os.unlink(path)
+            except OSError:
+                log.exception('failed to remove drained intent %s', path)
+
+        return 0
+    finally:
+        try:
+            fcntl.flock(lock_fd, fcntl.LOCK_UN)
+        finally:
+            os.close(lock_fd)
+
+
+if __name__ == '__main__':
+    sys.exit(main())
@@ -17,9 +17,9 @@ createrepo /nsm/repo
 # The kernel repo section is deployed to repodownload.conf by the manager highstate, which
 # runs AFTER this script during soup. On the first upgrade to a kernel-aware version the
 # on-disk config still predates the section, so guard on its presence to avoid dnf's
-# "Unknown repo: 'securityonionkernel'" aborting the sync (set -e). The next sync after the
+# "Unknown repo: 'securityonionkernelsync'" aborting the sync (set -e). The next sync after the
 # highstate deploys the section will pick it up.
-if grep -q '^\[securityonionkernel\]' /opt/so/conf/reposync/repodownload.conf; then
-  dnf reposync --norepopath -g --delete -m -c /opt/so/conf/reposync/repodownload.conf --repoid=securityonionkernel --download-metadata -p /nsm/kernelrepo/
+if grep -q '^\[securityonionkernelsync\]' /opt/so/conf/reposync/repodownload.conf; then
+  dnf reposync --norepopath -g --delete -m -c /opt/so/conf/reposync/repodownload.conf --repoid=securityonionkernelsync --download-metadata -p /nsm/kernelrepo/
  createrepo /nsm/kernelrepo
 fi
@@ -245,6 +245,7 @@ check_airgap() {
      UPDATE_DIR=/tmp/soagupdate/SecurityOnion
      AGDOCKER=/tmp/soagupdate/docker
      AGREPO=/tmp/soagupdate/minimal/Packages
+      AGUEKREPO=/tmp/soagupdate/uek/Packages
  else
      is_airgap=1
  fi
@@ -690,6 +691,21 @@ ensure_postgres_local_pillar() {
  chown -R socore:socore "$dir"
 }

+ensure_salt_local_pillar() {
+  # The salt.auto_apply settings (moved from global.push) are a new SOC settings
+  # module, so the new pillar/top.sls references salt.soc_salt / salt.adv_salt
+  # unconditionally. Managers upgrading from before this change have no
+  # /opt/so/saltstack/local/pillar/salt/ (make_some_dirs only runs at install
+  # time), so the stubs must be created here before salt-master restarts against
+  # the new top.sls.
+  echo "Ensuring salt local pillar stubs exist."
+  local dir=/opt/so/saltstack/local/pillar/salt
+  mkdir -p "$dir"
+  [[ -f "$dir/soc_salt.sls" ]] || touch "$dir/soc_salt.sls"
+  [[ -f "$dir/adv_salt.sls" ]] || touch "$dir/adv_salt.sls"
+  chown -R socore:socore "$dir"
+}
+
 ensure_postgres_secret() {
  # On a fresh install, generate_passwords + secrets_pillar seed
  # secrets:postgres_pass in /opt/so/saltstack/local/pillar/secrets.sls. That
@@ -873,6 +889,8 @@ update_kafka_metadata() {
 }

 up_to_3.2.0() {
+  ensure_salt_local_pillar
+
  fix_logstash_0013_lumberjack_pipeline_name

  pin_elasticsearch_data_retention_method
@@ -1004,13 +1022,19 @@ update_airgap_rules() {
  rsync -a $UPDATE_DIR/agrules/securityonion-resources/* /nsm/securityonion-resources/
 }

-update_airgap_repo() {
+update_airgap_repos() {
  # Update the files in the repo
-  echo "Syncing new updates to /nsm/repo"
-  rsync -a $AGREPO/* /nsm/repo/
-  echo "Creating repo"
+  echo "Syncing new updates to /nsm/repo & /nsm/kernelrepo"
+  # Airgap soup copies new files into the local repo, but doesn't remove old packages. Retaining the ability to rollback package updates
+  rsync -a "$AGREPO"/ /nsm/repo/
+  rsync -a "$AGUEKREPO"/ /nsm/kernelrepo/
+
  dnf -y install yum-utils createrepo_c
+
+  echo "Running createrepo for /nsm/repo"
  createrepo /nsm/repo
+  echo "Running createrepo for /nsm/kernelrepo"
+  createrepo /nsm/kernelrepo
 }

 update_salt_mine() {
@@ -1766,7 +1790,7 @@ main() {
  set -e

  if [[ $is_airgap -eq 0 ]]; then
-    update_airgap_repo
+    update_airgap_repos
    dnf clean all
    check_os_updates
  elif [[ $OS == 'oracle' ]]; then
@@ -34,6 +34,7 @@ make-rule-dir-nginx:
 so-nginx:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-nginx:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - hostname: so-nginx
    - networks:
      - sobridge:
@@ -0,0 +1,37 @@
+{% from 'salt/auto_apply.map.jinja' import AUTOAPPLY %}
+{% set actions = salt['pillar.get']('actions', []) %}
+{% set BATCH = AUTOAPPLY.batch %}
+{% set BATCH_WAIT = AUTOAPPLY.batch_wait %}
+
+{% for action in actions %}
+{%   if action.get('highstate') %}
+apply_highstate_{{ loop.index }}:
+  salt.state:
+    - tgt: '{{ action.tgt }}'
+    - tgt_type: {{ action.get('tgt_type', 'compound') }}
+    - highstate: True
+    - batch: {{ action.get('batch', BATCH) }}
+    - batch_wait: {{ action.get('batch_wait', BATCH_WAIT) }}
+    - kwarg:
+        queue: 2
+{%   else %}
+refresh_pillar_{{ loop.index }}:
+  salt.function:
+    - name: saltutil.refresh_pillar
+    - tgt: '{{ action.tgt }}'
+    - tgt_type: {{ action.get('tgt_type', 'compound') }}
+
+apply_{{ action.state | replace('.', '_') }}_{{ loop.index }}:
+  salt.state:
+    - tgt: '{{ action.tgt }}'
+    - tgt_type: {{ action.get('tgt_type', 'compound') }}
+    - sls:
+      - {{ action.state }}
+    - batch: {{ action.get('batch', BATCH) }}
+    - batch_wait: {{ action.get('batch_wait', BATCH_WAIT) }}
+    - kwarg:
+        queue: 2
+    - require:
+      - salt: refresh_pillar_{{ loop.index }}
+{%   endif %}
+{% endfor %}
@@ -0,0 +1,251 @@
+# One pillar directory can map to multiple (state, tgt) actions.
+# tgt is a raw salt compound expression. tgt_type is always "compound".
+# Per-action `batch` / `batch_wait` override the orch defaults (25% / 15s).
+# An action with `highstate: True` triggers state.highstate instead of
+# state.apply -- see salt/orch/push_batch.sls.
+#
+# Notes:
+#   - `bpf` is a pillar-only dir (no state of its own) consumed by both
+#     zeek and suricata via macros, so a bpf pillar change re-applies both.
+#   - suricata/strelka/zeek/elasticsearch/redis/kafka/logstash etc. have
+#     their own pillar dirs AND their own state, so they map 1:1 (or 1:2
+#     in strelka's case, because of the split init.sls / manager.sls).
+#
+# Intentional omissions (these will log a "not in pillar_push_map.yaml"
+# warning in push_pillar.sls and wait for the next scheduled highstate):
+#   - `data` and `node_data`: pillar-only data consumed by many states;
+#     handling them generically would amount to a fleetwide highstate.
+#   - `host`: soc_host describes mainint/mainip; a change is a re-IP and
+#     needs a coordinated procedure, not an immediate state push.
+#   - `hypervisor`: state changes touch libvirt and are disruptive; leave
+#     to the next scheduled highstate.
+#   - `sensor`: every field in soc_sensor.yaml is `readonly: True` or
+#     per-minion (`node: True`). Per-minion edits are persisted under
+#     pillar/minions/<id>.sls and are handled by Branch A of push_pillar.sls
+#     (per-minion highstate intent), not by this app-pillar map.
+#
+# The role sets here were verified line-by-line against salt/top.sls. If
+# salt/top.sls changes how an app is targeted, update the corresponding
+# compound here.
+
+# firewall: the one pillar everyone touches. Applied everywhere intentionally
+# because every host's iptables needs to know about every other host in the
+# grid. Salt's firewall state is idempotent (file.managed + iptables-restore
+# onchanges in salt/firewall/init.sls), so hosts whose rendered firewall is
+# unchanged do a file comparison and no-op without touching iptables -- actual
+# reload happens only on the hosts whose rules actually changed. Fleetwide
+# blast radius is intentional and matches the pre-plan behavior via highstate.
+# Adding N sensors in a burst coalesces into one dispatch via the drainer.
+firewall:
+  - state: firewall
+    tgt: '*'
+
+# backup: backup.config_backup runs on eval, standalone, manager, managerhype,
+# managersearch (NOT import -- the backup pillar is included on import per
+# pillar/top.sls but the backup state is not run there per salt/top.sls).
+backup:
+  - state: backup.config_backup
+    tgt: 'G@role:so-eval or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# bpf is pillar-only (no state); consumed by both zeek and suricata as macros.
+# Both states run on sensor_roles + so-import per salt/top.sls.
+bpf:
+  - state: zeek
+    tgt: 'G@role:so-eval or G@role:so-heavynode or G@role:so-import or G@role:so-sensor or G@role:so-standalone'
+  - state: suricata
+    tgt: 'G@role:so-eval or G@role:so-heavynode or G@role:so-import or G@role:so-sensor or G@role:so-standalone'
+
+# ca is applied universally.
+ca:
+  - state: ca
+    tgt: '*'
+
+# docker: universal. The docker state is in both the all-non-managers and
+# all-managers branches of salt/top.sls.
+docker:
+  - state: docker
+    tgt: '*'
+
+# elastalert: eval, standalone, manager, managerhype, managersearch (NOT import).
+elastalert:
+  - state: elastalert
+    tgt: 'G@role:so-eval or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# elastic-fleet-package-registry: manager_roles exactly.
+elastic-fleet-package-registry:
+  - state: elastic-fleet-package-registry
+    tgt: 'G@role:so-eval or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# elasticsearch: 8 roles.
+elasticsearch:
+  - state: elasticsearch
+    tgt: 'G@role:so-eval or G@role:so-heavynode or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-searchnode or G@role:so-standalone'
+
+# elasticagent: so-heavynode only.
+elasticagent:
+  - state: elasticagent
+    tgt: 'G@role:so-heavynode'
+
+# elasticfleet: base state only on pillar change. elasticfleet.install_agent_grid
+# is a deploy/enrollment step, not a config reload; leave it to the next highstate.
+elasticfleet:
+  - state: elasticfleet
+    tgt: 'G@role:so-eval or G@role:so-fleet or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# global: fanout to a fleetwide highstate. The global pillar (soc_global.sls)
+# carries cross-cutting settings (pipeline, url_base, imagerepo, mdengine, ...)
+# that are consumed by virtually every state, so a targeted re-apply isn't
+# meaningful. The drainer's batch/batch_wait throttling controls blast radius.
+global:
+  - highstate: True
+    tgt: '*'
+
+# healthcheck: eval, sensor, standalone only.
+healthcheck:
+  - state: healthcheck
+    tgt: 'G@role:so-eval or G@role:so-sensor or G@role:so-standalone'
+
+# hydra: manager_roles exactly.
+hydra:
+  - state: hydra
+    tgt: 'G@role:so-eval or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# idh: so-idh only.
+idh:
+  - state: idh
+    tgt: 'G@role:so-idh'
+
+# influxdb: manager_roles exactly.
+influxdb:
+  - state: influxdb
+    tgt: 'G@role:so-eval or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# kafka: standalone, manager, managerhype, managersearch, searchnode, receiver.
+kafka:
+  - state: kafka
+    tgt: 'G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-receiver or G@role:so-searchnode or G@role:so-standalone'
+
+# kibana: manager_roles exactly.
+kibana:
+  - state: kibana
+    tgt: 'G@role:so-eval or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# kratos: manager_roles exactly.
+kratos:
+  - state: kratos
+    tgt: 'G@role:so-eval or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# logrotate: universal (top-of-file '*' branch in salt/top.sls).
+logrotate:
+  - state: logrotate
+    tgt: '*'
+
+# logstash: 8 roles, no eval/import.
+logstash:
+  - state: logstash
+    tgt: 'G@role:so-fleet or G@role:so-heavynode or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-receiver or G@role:so-searchnode or G@role:so-standalone'
+
+# manager: manager_roles exactly. The manager state is also referenced under
+# *_sensor / *_heavynode top.sls blocks via `sensor`, but the standalone
+# `manager` state itself runs only on manager_roles.
+manager:
+  - state: manager
+    tgt: 'G@role:so-eval or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# nginx: 10 specific roles. NOT receiver, idh, hypervisor, desktop.
+nginx:
+  - state: nginx
+    tgt: 'G@role:so-eval or G@role:so-fleet or G@role:so-heavynode or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-searchnode or G@role:so-sensor or G@role:so-standalone'
+
+# ntp: universal (top-of-file '*' branch in salt/top.sls).
+ntp:
+  - state: ntp
+    tgt: '*'
+
+# patch: universal. soc_patch carries the OS update schedule, applied via
+# patch.os.schedule on every node (it's in both the all-non-managers and
+# all-managers branches of salt/top.sls).
+patch:
+  - state: patch.os.schedule
+    tgt: '*'
+
+# postgres: manager_roles exactly.
+postgres:
+  - state: postgres
+    tgt: 'G@role:so-eval or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# redis: 6 roles. standalone, manager, managerhype, managersearch, heavynode, receiver.
+# (NOT eval, NOT import, NOT searchnode.)
+redis:
+  - state: redis
+    tgt: 'G@role:so-heavynode or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-receiver or G@role:so-standalone'
+
+# registry: manager_roles exactly.
+registry:
+  - state: registry
+    tgt: 'G@role:so-eval or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# salt: fanout to a fleetwide highstate. The salt.auto_apply settings tune the
+# push pipeline itself (enabled, debounce/drain intervals, batch sizing) and
+# salt.schedule sets the per-minion highstate interval; they are consumed by the
+# manager's schedule, beacons, and master reactor config as well as every
+# minion's highstate schedule, so a targeted re-apply isn't meaningful. A salt
+# audit row only fires for SOC-driven salt.auto_apply / salt.schedule edits --
+# salt version bumps go through soup, not SOC, so they never reach this map.
+salt:
+  - highstate: True
+    tgt: '*'
+
+# sensoroni: universal.
+sensoroni:
+  - state: sensoroni
+    tgt: '*'
+
+# soc: manager_roles exactly.
+soc:
+  - state: soc
+    tgt: 'G@role:so-eval or G@role:so-import or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-standalone'
+
+# stig: broad. Runs on standalone, manager, managerhype, managersearch,
+# searchnode, sensor, receiver, fleet, hypervisor, desktop.
+# NOT eval, NOT import, NOT heavynode, NOT idh (the *_idh block in
+# salt/top.sls intentionally omits stig).
+stig:
+  - state: stig
+    tgt: 'G@role:so-desktop or G@role:so-fleet or G@role:so-hypervisor or G@role:so-manager or G@role:so-managerhype or G@role:so-managersearch or G@role:so-receiver or G@role:so-searchnode or G@role:so-sensor or G@role:so-standalone'
+
+# strelka: sensor-side only on pillar change (sensor_roles). strelka.manager is
+# intentionally NOT fired on pillar changes -- YARA rule and strelka config
+# pillar changes are consumed by the sensor-side strelka backend, and re-running
+# strelka.manager on managers is both unnecessary and disruptive. strelka.manager
+# is left to the 2-hour highstate.
+strelka:
+  - state: strelka
+    tgt: 'G@role:so-eval or G@role:so-heavynode or G@role:so-sensor or G@role:so-standalone'
+
+# suricata: sensor_roles + so-import (5 roles).
+suricata:
+  - state: suricata
+    tgt: 'G@role:so-eval or G@role:so-heavynode or G@role:so-import or G@role:so-sensor or G@role:so-standalone'
+
+# telegraf: universal.
+telegraf:
+  - state: telegraf
+    tgt: '*'
+
+# versionlock: universal (top-of-file '*' branch in salt/top.sls).
+versionlock:
+  - state: versionlock
+    tgt: '*'
+
+# vm: libvirt-driver hypervisors only. Matched by the salt-cloud:driver:libvirt
+# grain (compound supports nested grain matching via G@<key>:<subkey>:<value>).
+# pillar/vm/soc_vm.sls write path is referenced at salt/_runners/setup_hypervisor.py:856.
+vm:
+  - state: vm
+    tgt: 'G@salt-cloud:driver:libvirt'
+
+# zeek: sensor_roles + so-import (5 roles).
+zeek:
+  - state: zeek
+    tgt: 'G@role:so-eval or G@role:so-heavynode or G@role:so-import or G@role:so-sensor or G@role:so-standalone'
@@ -0,0 +1,176 @@
+#!py
+
+# Reactor invoked by the postgres_pillar_beacon when SOC records settings changes in
+# the securityonion.audit_settings table (see salt/_beacons/postgres_pillar_beacon.py). The beacon
+# emits one event per new row carrying setting_id and node_id.
+#
+# Two branches, keyed on node_id:
+#   A) node_id populated -> the change is scoped to that one minion. Look up the
+#      app in pillar_push_map.yaml and write an intent that runs the app's mapped
+#      state(s) targeted to just that node.
+#   B) node_id empty -> grid-wide app change. Look up the app in
+#      pillar_push_map.yaml and write an intent with the entry's actions as-is.
+#
+# The app name is the first dotted segment of setting_id (e.g. "telegraf.output"
+# -> "telegraf"), which matches the pillar_push_map.yaml keys 1:1.
+#
+# Reactors never dispatch directly. The so-push-drainer schedule picks up
+# ready intents, dedupes across pending files, and dispatches orch.push_batch.
+
+import fcntl
+import json
+import logging
+import os
+import time
+
+from salt.client import Caller
+import yaml
+
+LOG = logging.getLogger(__name__)
+
+PENDING_DIR = '/opt/so/state/push_pending'
+LOCK_FILE = os.path.join(PENDING_DIR, '.lock')
+MAX_PATHS = 20
+
+# The pillar_push_map.yaml is shipped via salt:// but the reactor runs on the
+# master, which mounts the default saltstack tree at this path.
+PUSH_MAP_PATH = '/opt/so/saltstack/default/salt/reactor/pillar_push_map.yaml'
+
+_PUSH_MAP_CACHE = {'mtime': 0, 'data': None}
+
+
+def _load_push_map():
+    try:
+        st = os.stat(PUSH_MAP_PATH)
+    except OSError:
+        LOG.warning('push_pillar: %s not found', PUSH_MAP_PATH)
+        return {}
+    if _PUSH_MAP_CACHE['mtime'] != st.st_mtime:
+        try:
+            with open(PUSH_MAP_PATH, 'r') as f:
+                _PUSH_MAP_CACHE['data'] = yaml.safe_load(f) or {}
+        except Exception:
+            LOG.exception('push_pillar: failed to load %s', PUSH_MAP_PATH)
+            _PUSH_MAP_CACHE['data'] = {}
+        _PUSH_MAP_CACHE['mtime'] = st.st_mtime
+    return _PUSH_MAP_CACHE['data'] or {}
+
+
+def _push_enabled():
+    try:
+        caller = Caller()
+        return bool(caller.cmd('pillar.get', 'salt:auto_apply:enabled', True))
+    except Exception:
+        LOG.exception('push_pillar: pillar.get salt:auto_apply:enabled failed, assuming enabled')
+        return True
+
+
+def _write_intent(key, actions, path):
+    now = time.time()
+    try:
+        os.makedirs(PENDING_DIR, exist_ok=True)
+    except OSError:
+        LOG.exception('push_pillar: cannot create %s', PENDING_DIR)
+        return
+
+    intent_path = os.path.join(PENDING_DIR, '{}.json'.format(key))
+    lock_fd = os.open(LOCK_FILE, os.O_CREAT | os.O_RDWR, 0o644)
+    try:
+        fcntl.flock(lock_fd, fcntl.LOCK_EX)
+
+        intent = {}
+        if os.path.exists(intent_path):
+            try:
+                with open(intent_path, 'r') as f:
+                    intent = json.load(f)
+            except (IOError, ValueError):
+                intent = {}
+
+        intent.setdefault('first_touch', now)
+        intent['last_touch'] = now
+        intent['actions'] = actions
+        paths = intent.get('paths', [])
+        if path and path not in paths:
+            paths.append(path)
+            paths = paths[-MAX_PATHS:]
+        intent['paths'] = paths
+
+        tmp_path = intent_path + '.tmp'
+        with open(tmp_path, 'w') as f:
+            json.dump(intent, f)
+        os.rename(tmp_path, intent_path)
+    except Exception:
+        LOG.exception('push_pillar: failed to write intent %s', intent_path)
+    finally:
+        try:
+            fcntl.flock(lock_fd, fcntl.LOCK_UN)
+        finally:
+            os.close(lock_fd)
+
+
+def _app_from_setting(setting_id):
+    # setting_id is e.g. 'telegraf.output' -> 'telegraf', 'ntp.config.servers' -> 'ntp'
+    if not setting_id:
+        return None
+    return setting_id.split('.', 1)[0] or None
+
+
+def _node_actions(entry, node_id):
+    # Copy the app's mapped actions but retarget each one to the single node.
+    # Preserves the state/highstate selection and any batch/batch_wait overrides.
+    actions = []
+    for action in entry:
+        if not isinstance(action, dict):
+            continue
+        node_action = dict(action)
+        node_action['tgt'] = node_id
+        node_action['tgt_type'] = 'glob'
+        actions.append(node_action)
+    return actions
+
+
+def run():
+    if not _push_enabled():
+        LOG.info('push_pillar: push disabled, skipping')
+        return {}
+
+    # The postgres_pillar_beacon nests its payload under data['data']; fall back to the
+    # top level so the reactor is robust to either shape.
+    event = data.get('data', data)  # noqa: F821 -- data provided by reactor
+    setting_id = event.get('setting_id', '')
+    node_id = (event.get('node_id') or '').strip()
+
+    app = _app_from_setting(setting_id)
+    if not app:
+        LOG.debug('push_pillar: ignoring event with no app segment: setting_id=%s', setting_id)
+        return {}
+
+    push_map = _load_push_map()
+    entry = push_map.get(app)
+    if not entry:
+        LOG.warning(
+            'push_pillar: app "%s" is not in pillar_push_map.yaml; change will be '
+            'picked up at the next scheduled highstate (setting_id=%s)',
+            app, setting_id,
+        )
+        return {}
+
+    # Branch A: per-node change -> retarget the app's states to just that node.
+    if node_id:
+        actions = _node_actions(entry, node_id)
+        if not actions:
+            LOG.warning('push_pillar: no usable actions for app "%s" (setting_id=%s)', app, setting_id)
+            return {}
+        _write_intent(
+            'node_{}_{}'.format(node_id, app), actions,
+            'audit:{}@{}'.format(setting_id, node_id),
+        )
+        LOG.info('push_pillar: per-node intent updated for %s on %s (setting_id=%s)',
+                 app, node_id, setting_id)
+        return {}
+
+    # Branch B: grid-wide app change -> use the map entry's actions as-is.
+    actions = list(entry)  # copy to avoid mutating the cache
+    _write_intent('pillar_{}'.format(app), actions, 'audit:{}'.format(setting_id))
+    LOG.info('push_pillar: app intent updated for %s (setting_id=%s)', app, setting_id)
+    return {}
@@ -0,0 +1,96 @@
+#!py
+
+# Reactor invoked by the rules_beacon poll beacon (salt/_beacons/rules_beacon.py) on rule
+# file changes under /opt/so/saltstack/local/salt/strelka/rules/compiled/.
+#
+# Writes (or updates) a push intent at /opt/so/state/push_pending/rules_strelka.json
+# and returns {}. The so-push-drainer schedule picks up ready intents, dedupes
+# across pending files, and dispatches orch.push_batch. Reactors never dispatch
+# directly -- see plan /home/mreeves/.claude/plans/goofy-marinating-hummingbird.md.
+
+import fcntl
+import json
+import logging
+import os
+import time
+
+from salt.client import Caller
+
+LOG = logging.getLogger(__name__)
+
+PENDING_DIR = '/opt/so/state/push_pending'
+LOCK_FILE = os.path.join(PENDING_DIR, '.lock')
+MAX_PATHS = 20
+
+# Mirrors GLOBALS.sensor_roles in salt/vars/globals.map.jinja. Sensor-side
+# strelka runs on exactly these four roles; so-import gets strelka.manager
+# instead, which is not fired on pillar changes.
+SENSOR_ROLES = ['so-eval', 'so-heavynode', 'so-sensor', 'so-standalone']
+
+
+def _sensor_compound():
+    return ' or '.join('G@role:{}'.format(r) for r in SENSOR_ROLES)
+
+
+def _push_enabled():
+    try:
+        caller = Caller()
+        return bool(caller.cmd('pillar.get', 'salt:auto_apply:enabled', True))
+    except Exception:
+        LOG.exception('push_strelka: pillar.get salt:auto_apply:enabled failed, assuming enabled')
+        return True
+
+
+def _write_intent(key, actions, path):
+    now = time.time()
+    try:
+        os.makedirs(PENDING_DIR, exist_ok=True)
+    except OSError:
+        LOG.exception('push_strelka: cannot create %s', PENDING_DIR)
+        return
+
+    intent_path = os.path.join(PENDING_DIR, '{}.json'.format(key))
+    lock_fd = os.open(LOCK_FILE, os.O_CREAT | os.O_RDWR, 0o644)
+    try:
+        fcntl.flock(lock_fd, fcntl.LOCK_EX)
+
+        intent = {}
+        if os.path.exists(intent_path):
+            try:
+                with open(intent_path, 'r') as f:
+                    intent = json.load(f)
+            except (IOError, ValueError):
+                intent = {}
+
+        intent.setdefault('first_touch', now)
+        intent['last_touch'] = now
+        intent['actions'] = actions
+        paths = intent.get('paths', [])
+        if path and path not in paths:
+            paths.append(path)
+            paths = paths[-MAX_PATHS:]
+        intent['paths'] = paths
+
+        tmp_path = intent_path + '.tmp'
+        with open(tmp_path, 'w') as f:
+            json.dump(intent, f)
+        os.rename(tmp_path, intent_path)
+    except Exception:
+        LOG.exception('push_strelka: failed to write intent %s', intent_path)
+    finally:
+        try:
+            fcntl.flock(lock_fd, fcntl.LOCK_UN)
+        finally:
+            os.close(lock_fd)
+
+
+def run():
+    if not _push_enabled():
+        LOG.info('push_strelka: push disabled, skipping')
+        return {}
+
+    path = data.get('path', '')  # noqa: F821 -- data provided by reactor
+    actions = [{'state': 'strelka', 'tgt': _sensor_compound()}]
+    _write_intent('rules_strelka', actions, path)
+    LOG.info('push_strelka: intent updated for path=%s', path)
+    return {}
@@ -0,0 +1,95 @@
+#!py
+
+# Reactor invoked by the rules_beacon poll beacon (salt/_beacons/rules_beacon.py) on rule
+# file changes under /opt/so/saltstack/local/salt/suricata/rules/.
+#
+# Writes (or updates) a push intent at /opt/so/state/push_pending/rules_suricata.json
+# and returns {}. The so-push-drainer schedule picks up ready intents, dedupes
+# across pending files, and dispatches orch.push_batch. Reactors never dispatch
+# directly -- see plan /home/mreeves/.claude/plans/goofy-marinating-hummingbird.md.
+
+import fcntl
+import json
+import logging
+import os
+import time
+
+from salt.client import Caller
+
+LOG = logging.getLogger(__name__)
+
+PENDING_DIR = '/opt/so/state/push_pending'
+LOCK_FILE = os.path.join(PENDING_DIR, '.lock')
+MAX_PATHS = 20
+
+# Mirrors GLOBALS.sensor_roles in salt/vars/globals.map.jinja. Suricata also
+# runs on so-import per salt/top.sls, so that role is appended below.
+SENSOR_ROLES = ['so-eval', 'so-heavynode', 'so-sensor', 'so-standalone']
+
+
+def _sensor_compound_plus_import():
+    return ' or '.join('G@role:{}'.format(r) for r in SENSOR_ROLES) + ' or G@role:so-import'
+
+
+def _push_enabled():
+    try:
+        caller = Caller()
+        return bool(caller.cmd('pillar.get', 'salt:auto_apply:enabled', True))
+    except Exception:
+        LOG.exception('push_suricata: pillar.get salt:auto_apply:enabled failed, assuming enabled')
+        return True
+
+
+def _write_intent(key, actions, path):
+    now = time.time()
+    try:
+        os.makedirs(PENDING_DIR, exist_ok=True)
+    except OSError:
+        LOG.exception('push_suricata: cannot create %s', PENDING_DIR)
+        return
+
+    intent_path = os.path.join(PENDING_DIR, '{}.json'.format(key))
+    lock_fd = os.open(LOCK_FILE, os.O_CREAT | os.O_RDWR, 0o644)
+    try:
+        fcntl.flock(lock_fd, fcntl.LOCK_EX)
+
+        intent = {}
+        if os.path.exists(intent_path):
+            try:
+                with open(intent_path, 'r') as f:
+                    intent = json.load(f)
+            except (IOError, ValueError):
+                intent = {}
+
+        intent.setdefault('first_touch', now)
+        intent['last_touch'] = now
+        intent['actions'] = actions
+        paths = intent.get('paths', [])
+        if path and path not in paths:
+            paths.append(path)
+            paths = paths[-MAX_PATHS:]
+        intent['paths'] = paths
+
+        tmp_path = intent_path + '.tmp'
+        with open(tmp_path, 'w') as f:
+            json.dump(intent, f)
+        os.rename(tmp_path, intent_path)
+    except Exception:
+        LOG.exception('push_suricata: failed to write intent %s', intent_path)
+    finally:
+        try:
+            fcntl.flock(lock_fd, fcntl.LOCK_UN)
+        finally:
+            os.close(lock_fd)
+
+
+def run():
+    if not _push_enabled():
+        LOG.info('push_suricata: push disabled, skipping')
+        return {}
+
+    path = data.get('path', '')  # noqa: F821 -- data provided by reactor
+    actions = [{'state': 'suricata', 'tgt': _sensor_compound_plus_import()}]
+    _write_intent('rules_suricata', actions, path)
+    LOG.info('push_suricata: intent updated for path=%s', path)
+    return {}
@@ -17,6 +17,7 @@ include:
 so-redis:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-redis:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - hostname: so-redis
    - user: socore
    - networks:
@@ -21,6 +21,9 @@ so-dockerregistry:
    - networks:
      - sobridge:
        - ipv4_address: {{ DOCKERMERGED.containers['so-dockerregistry'].ip }}
+    # Intentionally `always` (not unless-stopped) -- registry is critical infra
+    # and must come back up even if it was manually stopped. Do not homogenize
+    # to unless-stopped; see the container auto-restart section of the plan.
    - restart_policy: always
    - port_bindings:
      {% for BINDING in DOCKERMERGED.containers['so-dockerregistry'].port_bindings %}
@@ -0,0 +1,2 @@
+{% import_yaml 'salt/defaults.yaml' as SALT_DEFAULTS %}
+{% set AUTOAPPLY = salt['pillar.get']('salt:auto_apply', SALT_DEFAULTS.salt.auto_apply, merge=True) %}
@@ -3,7 +3,7 @@
 {% set SCHEDULE = salt['pillar.get']('healthcheck:schedule', 30) %}

 include:
-  - salt
+  - salt.minion

 {% if CHECKS and ENABLED %}
 salt_beacons:
@@ -14,12 +14,13 @@ salt_beacons:
    - defaults:
        CHECKS: {{ CHECKS }}
        SCHEDULE: {{ SCHEDULE }}
-    - watch_in: 
+    - watch_in:
      - service: salt_minion_service
 {% else %}
 salt_beacons:
  file.absent:
    - name: /etc/salt/minion.d/beacons.conf
-    - watch_in: 
+    - watch_in:
      - service: salt_minion_service
 {% endif %}
+
@@ -0,0 +1,9 @@
+salt:
+  auto_apply:
+    enabled: true
+    debounce_seconds: 30
+    drain_interval: 15
+    batch: '25%'
+    batch_wait: 15
+  schedule:
+    highstate_interval_hours: 2
@@ -0,0 +1,7 @@
+reactor:
+  - 'salt/beacon/*/rules_beacon/suricata':
+    - salt://reactor/push_suricata.sls
+  - 'salt/beacon/*/rules_beacon/strelka':
+    - salt://reactor/push_strelka.sls
+  - 'salt/beacon/*/postgres_pillar_beacon/audit_settings':
+    - salt://reactor/push_pillar.sls
@@ -0,0 +1,11 @@
+{% from 'vars/globals.map.jinja' import GLOBALS %}
+{% from 'salt/schedule.map.jinja' import SCHEDULEMERGED %}
+
+highstate_schedule:
+  schedule.present:
+    - function: state.highstate
+    - hours: {{ SCHEDULEMERGED.highstate_interval_hours }}
+    - maxrunning: 1
+{% if not GLOBALS.is_manager %}
+    - splay: 1800
+{% endif %}
@@ -5,3 +5,11 @@ salt_bootstrap:
    - source: salt://salt/scripts/bootstrap-salt.sh
    - mode: 755
    - show_changes: False
+
+salt_sbin:
+  file.recurse:
+    - name: /usr/sbin
+    - source: salt://salt/tools/sbin
+    - user: 939
+    - group: 939
+    - file_mode: 755
@@ -1,4 +1,4 @@
 lasthighstate:
  file.touch:
    - name: /opt/so/log/salt/lasthighstate
-    - order: last
+    - order: 9001
@@ -10,10 +10,12 @@
 #    software that is protected by the license key."

 {% from 'allowed_states.map.jinja' import allowed_states %}
+{% from 'salt/auto_apply.map.jinja' import AUTOAPPLY %}
 {% if sls in allowed_states %}

 include:
  - salt.minion
+  - salt.master.pyinotify
  - salt.master.boot_mine_update
 {%   if 'vrt' in salt['pillar.get']('features', []) %}
  - salt.cloud
@@ -63,6 +65,21 @@ engines_config:
    - name: /etc/salt/master.d/engines.conf
    - source: salt://salt/files/engines.conf

+{% if AUTOAPPLY.enabled %}
+reactor_pushstate_config:
+  file.managed:
+    - name: /etc/salt/master.d/reactor_pushstate.conf
+    - source: salt://salt/files/reactor_pushstate.conf
+    - watch_in:
+      - service: salt_master_service
+{% else %}
+reactor_pushstate_config:
+  file.absent:
+    - name: /etc/salt/master.d/reactor_pushstate.conf
+    - watch_in:
+      - service: salt_master_service
+{% endif %}
+
 # update the bootstrap script when used for salt-cloud
 salt_bootstrap_cloud:
  file.managed:
@@ -78,7 +95,7 @@ salt_master_service:
      - file: checkmine_engine
      - file: pillarWatch_engine
      - file: engines_config
-    - order: last
+    - order: 9002

 {% else %}

@@ -0,0 +1,20 @@
+# Copyright Security Onion Solutions LLC and/or licensed to Security Onion Solutions LLC under one
+# or more contributor license agreements. Licensed under the Elastic License 2.0 as shown at
+# https://securityonion.net/license; you may not use this file except in compliance with the
+# Elastic License 2.0.
+
+pyinotify_module_package:
+  file.recurse:
+    - name: /opt/so/conf/salt/module_packages/pyinotify
+    - source: salt://salt/module_packages/pyinotify
+    - clean: True
+    - makedirs: True
+
+pyinotify_python_module_install:
+  cmd.run:
+    - name: /opt/saltstack/salt/bin/python3.10 -m pip install pyinotify --no-index --find-links=/opt/so/conf/salt/module_packages/pyinotify/ --upgrade
+    - onchanges:
+      - file: pyinotify_module_package
+    - failhard: True
+    - watch_in:
+      - service: salt_minion_service
@@ -2,4 +2,3 @@
 salt:
  minion:
    version: '3006.19'
-    check_threshold: 3600 # in seconds, threshold used for so-salt-minion-check. any value less than 600 seconds may cause a lot of salt-minion restarts since the job to touch the file occurs every 5-8 minutes by default
@@ -111,13 +111,17 @@ mark_setup_complete_for_upgrades:

 {% endif %}

-# this has to be outside the if statement above since there are <requisite>_in calls to this state
+# this has to be outside the if statement above since there are <requisite>_in calls to this state.
+# uses watch (not listen) so the restart fires in-state and its result lands on this state's
+# running entry; that is what lets wait_for_salt_minion_ready below detect any restart
+# uniformly via onchanges, regardless of whether the trigger came from these files or from
+# external watch_in's (e.g. beacons, master/pyinotify).
 salt_minion_service:
  service.running:
    - name: salt-minion
    - enable: True
    - onlyif: test "{{INSTALLEDSALTVERSION}}" == "{{SALTVERSION}}"
-    - listen:
+    - watch:
      - file: mine_functions
 {% if INSTALLEDSALTVERSION|string == SALTVERSION|string %}
      - file: set_log_levels
@@ -126,3 +130,17 @@ salt_minion_service:
      - file: signing_policy
 {% endif %}
    - order: last
+
+# block until the just-restarted salt-minion is back and can execute modules locally, so
+# follow-on jobs and the next highstate iteration do not race the restart. onchanges +
+# require on salt_minion_service catches every restart trigger uniformly because watch
+# mod_watch results replace the service state's running entry. wait logic lives in
+# /usr/sbin/so-salt-minion-wait (deployed by common_sbin from common/tools/sbin/).
+wait_for_salt_minion_ready:
+  cmd.run:
+    - name: /usr/sbin/so-salt-minion-wait
+    - onchanges:
+      - service: salt_minion_service
+    - require:
+      - service: salt_minion_service
+    - order: last
@@ -0,0 +1,17 @@
+{% from 'vars/globals.map.jinja' import GLOBALS %}
+{% from 'salt/auto_apply.map.jinja' import AUTOAPPLY %}
+
+{% if GLOBALS.is_manager and AUTOAPPLY.enabled %}
+push_drain_schedule:
+  schedule.present:
+    - function: cmd.run
+    - job_args:
+      - /usr/sbin/so-push-drainer
+    - seconds: {{ AUTOAPPLY.drain_interval }}
+    - maxrunning: 1
+    - return_job: False
+{% elif GLOBALS.is_manager %}
+push_drain_schedule:
+  schedule.absent:
+    - name: push_drain_schedule
+{% endif %}
@@ -0,0 +1,2 @@
+{% import_yaml 'salt/defaults.yaml' as SALT_DEFAULTS %}
+{% set SCHEDULEMERGED = salt['pillar.get']('salt:schedule', SALT_DEFAULTS.salt.schedule, merge=True) %}
@@ -0,0 +1,39 @@
+salt:
+  auto_apply:
+    enabled:
+      description: Master kill-switch for the active push feature. When disabled, rule and pillar changes are picked up at the next scheduled highstate instead of being pushed immediately.
+      forcedType: bool
+      helpLink: push
+      global: True
+    debounce_seconds:
+      description: Trailing-edge debounce window in seconds. A push intent must be quiet for this long before the drainer dispatches. Rapid bursts of edits within this window coalesce into one dispatch.
+      forcedType: int
+      helpLink: push
+      global: True
+      advanced: True
+    drain_interval:
+      description: How often the push drainer checks for ready intents, in seconds. Small values lower dispatch latency at the cost of more background work on the manager.
+      forcedType: int
+      helpLink: push
+      global: True
+      advanced: True
+    batch:
+      description: "Host batch size for push orchestrations. A number (e.g. '10') or a percentage (e.g. '25%'). Limits how many minions run the push state at once so large fleets don't thundering-herd."
+      helpLink: push
+      global: True
+      advanced: True
+      regex: '^([0-9]+%?)$'
+      regexFailureMessage: Enter a whole number or a whole-number percentage (e.g. 10 or 25%).
+    batch_wait:
+      description: Seconds to wait between host batches in a push orchestration. Gives the fleet time to breathe between waves.
+      forcedType: int
+      helpLink: push
+      global: True
+      advanced: True
+  schedule:
+    highstate_interval_hours:
+      description: How often every minion in the grid runs a scheduled state.highstate, in hours. Lower values keep minions closer in sync at the cost of more load; higher values reduce load but increase worst-case latency for non-pushed changes. The salt-minion health check restarts a minion if its last highstate is older than this value plus one hour.
+      forcedType: int
+      helpLink: push
+      global: True
+      advanced: True
@@ -0,0 +1,35 @@
+#!/bin/bash
+#
+# Copyright Security Onion Solutions LLC and/or licensed to Security Onion Solutions LLC under one
+# or more contributor license agreements. Licensed under the Elastic License 2.0 as shown at
+# https://securityonion.net/license; you may not use this file except in compliance with the
+# Elastic License 2.0.
+
+# Block until the local salt-minion service is back up and can execute modules locally.
+# Invoked from the wait_for_salt_minion_ready state in salt/minion/init.sls after
+# salt_minion_service fires its watch-driven mod_watch (a non-blocking systemctl restart),
+# so follow-on jobs and the next highstate iteration do not race the in-flight restart.
+
+. /usr/sbin/so-common
+
+# Initial sleep gives the systemctl restart (--no-block by default for salt-minion on
+# >=3006.15) time to begin tearing down the old process before we probe for readiness.
+INITIAL_SLEEP=3
+TIMEOUT=120
+PING_TIMEOUT=5
+
+sleep "$INITIAL_SLEEP"
+
+elapsed="$INITIAL_SLEEP"
+while [ "$elapsed" -lt "$TIMEOUT" ]; do
+  if systemctl is-active --quiet salt-minion \
+     && salt-call --local --timeout="$PING_TIMEOUT" --out=quiet test.ping >/dev/null 2>&1; then
+    echo "salt-minion ready after ${elapsed}s"
+    exit 0
+  fi
+  sleep 1
+  elapsed=$((elapsed + 1))
+done
+
+echo "salt-minion did not become ready within ${TIMEOUT}s" >&2
+exit 1
@@ -1,10 +0,0 @@
-{%   from 'vars/globals.map.jinja' import GLOBALS %}
-
-highstate_schedule:
-  schedule.present:
-    - function: state.highstate
-    - minutes: 15
-    - maxrunning: 1
-{% if not GLOBALS.is_manager %}
-    - splay: 120
-{% endif %}
@@ -14,6 +14,7 @@ include:
 so-sensoroni:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-soc:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - network_mode: host
    - binds:
      - /nsm/import:/nsm/import:rw
@@ -18,6 +18,7 @@ include:
 so-soc:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-soc:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - hostname: soc
    - name: so-soc
    - networks:
@@ -47,6 +47,10 @@ strelka_backend:
      - {{ ULIMIT.name }}={{ ULIMIT.soft }}:{{ ULIMIT.hard }}
    {%   endfor %}
    {% endif %}
+    # Intentionally `on-failure` (not unless-stopped) -- strelka backend shuts
+    # down cleanly during rule reloads and we do not want those clean exits to
+    # trigger an auto-restart. Do not homogenize; see the container
+    # auto-restart section of the plan.
    - restart_policy: on-failure
    - watch:
      - file: strelkasensorcompiledrules
@@ -15,6 +15,7 @@ include:
 strelka_coordinator:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-redis:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - name: so-strelka-coordinator
    - networks:
      - sobridge:
@@ -15,6 +15,7 @@ include:
 strelka_filestream:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-strelka-manager:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - binds:
      - /opt/so/conf/strelka/filestream/:/etc/strelka/:ro
      - /nsm/strelka:/nsm/strelka
@@ -15,6 +15,7 @@ include:
 strelka_frontend:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-strelka-manager:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - binds:
      - /opt/so/conf/strelka/frontend/:/etc/strelka/:ro
      - /nsm/strelka/log/:/var/log/strelka/:rw
@@ -15,6 +15,7 @@ include:
 strelka_gatekeeper:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-redis:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - name: so-strelka-gatekeeper
    - networks:
      - sobridge:
@@ -15,6 +15,7 @@ include:
 strelka_manager:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-strelka-manager:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - binds:
      - /opt/so/conf/strelka/manager/:/etc/strelka/:ro
      {% if DOCKERMERGED.containers['so-strelka-manager'].custom_bind_mounts %}
@@ -18,6 +18,7 @@ so-suricata:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-suricata:{{ GLOBALS.so_version }}
    - privileged: True
+    - restart_policy: unless-stopped
    - environment:
      - INTERFACE={{ GLOBALS.sensor.interface }}
      {% if DOCKERMERGED.containers['so-suricata'].extra_env %}
@@ -65,10 +66,11 @@ so-suricata:
      - file: suriclassifications

 surirulereload:
-  cmd.run: 
+  cmd.run:
    - name: /usr/sbin/so-suricata-reload-rules >> /opt/so/log/suricata/reload.log 2>&1
-    - onchanges: 
+    - onchanges:
        - file: surirulesync
+    - onlyif: test -f /opt/so/rules/suricata/all-rulesets.rules
    - require:
        - docker_container: so-suricata
    
@@ -7,5 +7,59 @@

 . /usr/sbin/so-common

-retry 60 3 'docker exec so-suricata /opt/suricata/bin/suricatasc -c reload-rules /var/run/suricata/suricata-command.socket' '{"message":"done","return":"OK"}' || fail "The Suricata container was not ready in time."
-retry 60 3 'docker exec so-suricata /opt/suricata/bin/suricatasc -c ruleset-reload-nonblocking /var/run/suricata/suricata-command.socket' '{"message":"done","return":"OK"}' || fail "The Suricata container was not ready in time."
+RULES_FILE="/opt/so/rules/suricata/all-rulesets.rules"
+SOCKET="/var/run/suricata/suricata-command.socket"
+SURICATASC="docker exec so-suricata /opt/suricata/bin/suricatasc"
+
+# Format an epoch as a human-readable local timestamp for log messages.
+fmt_time() { date -d "@$1" '+%Y-%m-%d %H:%M:%S %Z' 2>/dev/null; }
+
+# Prefix each input line with the current timestamp.
+timestamp_lines() { while IFS= read -r line; do printf '%s %s\n' "$(date '+%Y-%m-%d %H:%M:%S %Z')" "$line"; done; }
+
+# Epoch of Suricata's last *completed* ruleset reload; non-zero return on failure.
+suricata_reload_epoch() {
+  local out ts
+  out=$($SURICATASC -c ruleset-reload-time "$SOCKET" 2>/dev/null)
+  ts=$(echo "$out" | jq -r '.message[0].last_reload // empty' 2>/dev/null)
+  [ -n "$ts" ] || return 1
+  date -d "$ts" +%s 2>/dev/null
+}
+
+# Trigger a fresh reload and confirm Suricata is running a ruleset at least as new
+# as the rules file. Returns 0 only when both hold, so retry keeps going until an
+# in-progress reload clears and our own reload completes.
+reload_and_verify() {
+  local out reload_epoch
+  out=$($SURICATASC -c reload-rules "$SOCKET")
+  echo "reload-rules: $out"
+
+  if [[ "$out" =~ "Reload already in progress" ]]; then
+    echo "A reload is already in progress; waiting for it to clear so a fresh reload can load the current ruleset."
+    return 1
+  fi
+  if [[ ! "$out" =~ '{"message":"done","return":"OK"}' ]]; then
+    echo "Suricata not ready or unexpected reload output; will retry."
+    return 1
+  fi
+
+  reload_epoch=$(suricata_reload_epoch) || { echo "Could not read ruleset-reload-time; will retry."; return 1; }
+  if [ "$reload_epoch" -ge "$target_mtime" ]; then
+    echo "Loaded ruleset is current: last reload ($(fmt_time "$reload_epoch")) is newer than rules file ($(fmt_time "$target_mtime"))."
+    return 0
+  fi
+  echo "Loaded ruleset is stale: last reload ($(fmt_time "$reload_epoch")) is older than rules file ($(fmt_time "$target_mtime")); retrying."
+  return 1
+}
+
+# Run the reload/verify, timestamping every line of output (ours and the
+# retry/fail helpers') so reload.log shows when each step ran. The pipeline is
+# synchronous, so the log is fully flushed and ordered before we exit; the
+# script's real exit code is preserved via PIPESTATUS.
+{
+  # Epoch mtime of the ruleset we need Suricata to have loaded. Captured once so
+  # a file update mid-reload does not move the goalpost.
+  target_mtime=$(stat -c %Y "$RULES_FILE") || fail "Could not stat the Suricata rules file: $RULES_FILE"
+  retry 60 3 'reload_and_verify' || fail "Suricata did not load the current ruleset in time."
+} 2>&1 | timestamp_lines
+exit "${PIPESTATUS[0]}"
@@ -7,6 +7,7 @@ so-tcpreplay:
  docker_container.running:
    - network_mode: "host"
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-tcpreplay:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - name: so-tcpreplay
    - user: root
    - interactive: True
@@ -18,6 +18,7 @@ include:
 so-telegraf:
  docker_container.running:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-telegraf:{{ GLOBALS.so_version }}
+    - restart_policy: unless-stopped
    - user: 939
    - group_add: 939,920
    - environment:
@@ -19,7 +19,7 @@ base:
    - repo.client
    - versionlock
    - ntp
-    - schedule
+    - salt.highstate_schedule
    - logrotate

  # manager node on proper salt version with empty node_data pillar
@@ -55,6 +55,7 @@ base:
    - motd
    - salt.minion-check
    - salt.lasthighstate
+    - salt.push_drain_schedule
    - common
    - docker
    - docker_clean
@@ -83,7 +84,6 @@ base:
    - zeek
    - strelka
    - elastalert
-    - utility
    - elasticfleet
    - pcap.cleanup

@@ -113,7 +113,6 @@ base:
    - zeek
    - strelka
    - elastalert
-    - utility
    - elasticfleet
    - stig
    - kafka
@@ -141,7 +140,6 @@ base:
    - elastic-fleet-package-registry
    - kibana
    - elastalert
-    - utility
    - elasticfleet
    - stig
    - kafka
@@ -168,7 +166,6 @@ base:
    - elastic-fleet-package-registry
    - kibana
    - elastalert
-    - utility
    - elasticfleet
    - kafka

@@ -198,7 +195,6 @@ base:
    - elastic-fleet-package-registry
    - kibana
    - elastalert
-    - utility
    - elasticfleet
    - stig
    - kafka
@@ -222,7 +218,6 @@ base:
    - elasticsearch
    - elastic-fleet-package-registry
    - kibana
-    - utility
    - suricata
    - zeek
    - elasticfleet
@@ -300,7 +295,6 @@ base:
    - nginx
    - elasticfleet
    - elasticfleet.install_agent_grid
-    - schedule
    - stig

  '*_hypervisor and I@features:vrt and G@saltversion:{{saltversion}}':
@@ -1,29 +0,0 @@
-#!/bin/bash
-# Wait for ElasticSearch to come up, so that we can query for version infromation
-echo -n "Waiting for ElasticSearch..."
-COUNT=0
-ELASTICSEARCH_CONNECTED="no"
-while [[ "$COUNT" -le 30 ]]; do
-  curl -K /opt/so/conf/elasticsearch/curl.config -k --output /dev/null --silent --head --fail -L https://{{ GLOBALS.manager_ip }}:9200
-  if [ $? -eq 0 ]; then
-    ELASTICSEARCH_CONNECTED="yes"
-    echo "connected!"
-    break
-  else
-    ((COUNT+=1))
-    sleep 1
-    echo -n "."
-  fi
-done
-if [ "$ELASTICSEARCH_CONNECTED" == "no" ]; then
-  echo
-  echo -e "Connection attempt timed out.  Unable to connect to ElasticSearch.  \nPlease try: \n  -checking log(s) in /var/log/elasticsearch/\n  -running 'docker ps' \n  -running 'sudo so-elastic-restart'"
-  echo
-
-  exit
-fi
-
-echo "Applying cross cluster search config..."
-    curl -K /opt/so/conf/elasticsearch/curl.config -s -k -XPUT -L https://{{ GLOBALS.manager_ip }}:9200/_cluster/settings \
-         -H 'Content-Type: application/json' \
-         -d "{\"persistent\": {\"search\": {\"remote\": {\"{{ grains.host }}\": {\"seeds\": [\"127.0.0.1:9300\"]}}}}}"
@@ -1,22 +0,0 @@
-{% from 'allowed_states.map.jinja' import allowed_states %}
-{% from 'vars/globals.map.jinja' import GLOBALS %}
-
-{% if sls in allowed_states %}
-  {% if grains['role'] in ['so-eval', 'so-import'] %}
-fixsearch:
-  cmd.script:
-    - shell: /bin/bash
-    - cwd: /opt/so
-    - source: salt://utility/bin/eval
-    - template: jinja
-    - defaults:
-        GLOBALS: {{ GLOBALS }}
-  {% endif %}
-
-{% else %}
-
-{{sls}}_state_not_allowed:
-  test.fail_without_changes:
-    - name: {{sls}}_state_not_allowed
-
-{% endif %}
@@ -18,6 +18,7 @@ so-zeek:
    - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-zeek:{{ GLOBALS.so_version }}
    - start: True
    - privileged: True
+    - restart_policy: unless-stopped
    {% if DOCKERMERGED.containers['so-zeek'].ulimits %}
    - ulimits:
    {%   for ULIMIT in DOCKERMERGED.containers['so-zeek'].ulimits %}
@@ -29,8 +29,12 @@ title() {
 }

 fail_setup() {
+	local err_msg=$1
+	if [[ -n "$err_msg" ]]; then
+		error "$err_msg"
+	fi
 	error "Setup encountered an unrecoverable failure, exiting"
-	touch /root/failure
+	echo "setup incomplete: $err_msg" > /root/failure
 	exit 1
 }

@@ -697,7 +701,7 @@ compare_main_nic_ip() {
 			EOM

 		  [[ -n $TESTING ]] || whiptail --title "$whiptail_title" --msgbox "$message" 11 75
-		  kill -SIGINT "$(ps --pid $$ -oppid=)"; fail_setup
+		  kill -SIGINT "$(ps --pid $$ -oppid=)"; fail_setup "Main IP mismatch"
 	  fi
 	else
 		# Setup uses MAINIP, but since we ignore the equality condition when using a VPN 
@@ -755,8 +759,7 @@ configure_management_bond() {
 	info "Setting up $bond_name management interface with mode $bond_mode"

 	if [[ ${#MBNICS[@]} -eq 0 ]]; then
-		error "[ERROR] No management bond NICs were selected."
-		fail_setup
+		fail_setup "No management bond NICs selected"
 	fi

 	nmcli -t -f NAME con show | grep -Fxq "$bond_name"
@@ -914,8 +917,7 @@ detect_os() {
 		is_rpm=true
 		is_supported=true
 	else
-		info "This OS is not supported. Security Onion requires Oracle Linux 9."
-		fail_setup
+		fail_setup "This OS is not supported. Security Onion requires Oracle Linux 9."
 	fi

 	info "Found OS: $OS $OSVER"
@@ -923,7 +925,7 @@ detect_os() {

 download_elastic_agent_artifacts() {
 	if ! update_elastic_agent 2>&1 | tee -a "$setup_log"; then
-		fail_setup
+		fail_setup "Failed to update Elastic Agent"
 	fi
 }

@@ -1433,7 +1435,7 @@ make_some_dirs() {
 	mkdir -p $local_salt_dir/salt/firewall/portgroups
 	mkdir -p $local_salt_dir/salt/firewall/ports

-	for THEDIR in bpf elasticsearch ntp firewall redis backup influxdb postgres strelka sensoroni soc docker zeek suricata nginx telegraf logstash soc manager kratos hydra idh elastalert stig global kafka versionlock hypervisor vm; do
+	for THEDIR in bpf elasticsearch ntp firewall redis backup influxdb postgres strelka sensoroni soc docker zeek suricata nginx telegraf logstash soc manager kratos hydra idh elastalert stig global salt kafka versionlock hypervisor vm; do
 	mkdir -p $local_salt_dir/pillar/$THEDIR
 	touch $local_salt_dir/pillar/$THEDIR/adv_$THEDIR.sls
 	touch $local_salt_dir/pillar/$THEDIR/soc_$THEDIR.sls
@@ -1567,7 +1569,7 @@ proxy_validate() {
 		error "Received error: $proxy_test_err"
 		if [[ -n $TESTING ]]; then
 			error "Exiting setup"
-			kill -SIGINT "$(ps --pid $$ -oppid=)"; fail_setup
+			kill -SIGINT "$(ps --pid $$ -oppid=)"; fail_setup "Proxy validation failed"
 		fi
 	fi
 	return $ret
@@ -1774,8 +1776,7 @@ ensure_pyyaml() {
 	local result=$?
 	set +o pipefail
 	if [[ $result -ne 0 ]] || ! rpm -q python3-pyyaml >/dev/null 2>&1; then
-		error "Failed to install python3-pyyaml (exit=$result)"
-		fail_setup
+		fail_setup "Failed to install python3-pyyaml (exit=$result)"
 	fi
 	info "python3-pyyaml installed successfully"
 }
@@ -1910,8 +1911,8 @@ repo_sync_local() {

 	if [[ ! $is_airgap ]]; then
 		curl --retry 5 --retry-delay 60 -A "netinstall/$SOVERSION/$OS/$(uname -r)/1" https://sigs.securityonion.net/checkup --output /tmp/install
-		retry 5 60 "dnf reposync --norepopath -g --delete -m -c /opt/so/conf/reposync/repodownload.conf --repoid=securityonionsync --download-metadata -p /nsm/repo/" >> "$setup_log" 2>&1 || fail_setup
-		retry 5 60 "dnf reposync --norepopath -g --delete -m -c /opt/so/conf/reposync/repodownload.conf --repoid=securityonionkernel --download-metadata -p /nsm/kernelrepo/" >> "$setup_log" 2>&1 || fail_setup
+		retry 5 60 "dnf reposync --norepopath -g --delete -m -c /opt/so/conf/reposync/repodownload.conf --repoid=securityonionsync --download-metadata -p /nsm/repo/" >> "$setup_log" 2>&1 || fail_setup "Failed to sync repos"
+		retry 5 60 "dnf reposync --norepopath -g --delete -m -c /opt/so/conf/reposync/repodownload.conf --repoid=securityonionkernel --download-metadata -p /nsm/kernelrepo/" >> "$setup_log" 2>&1 || fail_setup "Failed to sync kernel repos"
 		# After the download is complete run createrepo
 		create_repo
 	fi
@@ -1924,10 +1925,10 @@ saltify() {

 	if [[ $waitforstate ]]; then
 		# install all for a manager
-		retry 30 10 "bash ../salt/salt/scripts/bootstrap-salt.sh -r -M -X stable $SALTVERSION" || fail_setup
+		retry 30 10 "bash ../salt/salt/scripts/bootstrap-salt.sh -r -M -X stable $SALTVERSION" || fail_setup "Failed to install salt master"
 	else
 		# just a minion
-		retry 30 10 "bash ../salt/salt/scripts/bootstrap-salt.sh -r -X stable $SALTVERSION" || fail_setup
+		retry 30 10 "bash ../salt/salt/scripts/bootstrap-salt.sh -r -X stable $SALTVERSION" || fail_setup "Failed to install salt minion"
 	fi

 	salt_install_module_deps
@@ -1999,7 +2000,7 @@ set_main_ip() {
 			info "MAINIP=$MAINIP"
 			info "MNIC_IP=$MNIC_IP"
 			whiptail_error_message "The management IP could not be determined. Please check the log at /root/sosetup.log and verify the network configuration. Select OK to exit."
-			fail_setup
+			fail_setup "Could not determine MAINIP or MNIC_IP"
 		fi
 		sleep 1
 	done
@@ -2203,7 +2204,7 @@ set_initial_firewall_access() {
 set_management_interface() {
 	title "Setting up the main interface"
 	if [[ $MNIC == "bond1" ]]; then
-		configure_management_bond || fail_setup
+		configure_management_bond || fail_setup "Failed to configure management bond"
 	fi

 	if [ "$address_type" = 'DHCP' ]; then
@@ -90,8 +90,7 @@ if [[ "$setup_type" == 'iso' ]]; then
 	if [[ $is_rpm ]]; then
 		is_iso=true
 	else
-		echo "Only use 'so-setup iso' for an ISO install on Security Onion ISO images. Please run 'so-setup network' instead."
-		fail_setup
+		fail_setup "Only use 'so-setup iso' for an ISO install on Security Onion ISO images. Please run 'so-setup network' instead."
 	fi
 fi

@@ -130,7 +129,7 @@ catch() {
 	info "Fatal error occurred at $1 in so-setup, failing setup."
 	grep --color=never "ERROR" "$setup_log" > "$error_log"
 	whiptail_setup_failed
-	fail_setup
+	fail_setup "Fatal error occurred at $1 in so-setup"
 }

 # Add the progress function for manager node type installs
@@ -238,8 +237,7 @@ case "$setup_type" in
 		info "Beginning Security Onion $setup_type install"
 		;;
 	*)
-		error "Invalid install type, must be 'iso', 'network' or 'desktop'."
-		fail_setup
+		fail_setup "Invalid install type, must be 'iso', 'network' or 'desktop'."
 		;;
 esac

@@ -773,8 +771,7 @@ if ! [[ -f $install_opt_file ]]; then
 		logCmd "salt-call state.apply -l info registry"
 		title "Seeding the docker registry"
 		if ! docker_seed_registry; then
-			error "Failed to seed the docker registry"
-			fail_setup
+			fail_setup "Failed to seed the docker registry"
 		fi
 		title "Applying the manager state"
 		logCmd "salt-call state.apply -l info manager"
@@ -797,8 +794,7 @@ if ! [[ -f $install_opt_file ]]; then
 		title "Setting up Elastic Fleet"
 		logCmd "salt-call state.apply elasticfleet.config"	
 		if ! logCmd so-elastic-fleet-setup; then
-			error "Failed to run so-elastic-fleet-setup"
-			fail_setup
+			fail_setup "Failed to run so-elastic-fleet-setup"
 		fi
 		mark_setup_complete
 		set_initial_firewall_access
@@ -143,15 +143,15 @@ main() {
        cat $error_log
        echo "--------------------------"
        exit_code=1
-        touch /root/failure
+        echo "Found setup errors. Check $error_log for details" > /root/failure
    elif using_iso && cron_error_in_mail_spool; then
        echo "WARNING: Unexpected cron job output in mail spool"
        exit_code=1
-        touch /root/failure
+        echo "Unexpected cron job output found in /var/spool/mail/" > /root/failure
    elif is_manager_node && status_failed; then
        echo "WARNING: Containers are not in a healthy state"
        exit_code=1
-        touch /root/failure
+        echo "Containers are not in a healthy state. Check so-status for details" > /root/failure
    else
        echo "Successfully completed setup!"
        touch /root/success
Author	SHA1	Message	Date
Josh Patterson	3310e19ee4	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-07-02 10:27:54 -04:00
Jason Ertel	07d6b2cfdd	Merge pull request #16033 from Security-Onion-Solutions/jertel/wip avoid setup failure reason ambiguity	2026-07-02 09:20:48 -04:00
Jason Ertel	89afea876a	Merge branch '3/dev' into jertel/wip	2026-07-02 09:04:57 -04:00
Jason Ertel	1243a25bd3	avoid setup failure reason ambiguity	2026-07-02 08:59:52 -04:00
Josh Patterson	76f6947f36	Merge pull request #16029 from Security-Onion-Solutions/surirulereload only reload suricata rules if all-rulesets.rules exists	2026-07-01 16:54:02 -04:00
Jorge Reyes	92a55386c6	Merge pull request #16028 from Security-Onion-Solutions/reyesj2-patch-6 duplicate repo name in so-repo-sync	2026-07-01 15:50:54 -05:00
reyesj2	e7352eb841	duplicate repo name in so-repo-sync	2026-07-01 15:17:55 -05:00
Josh Patterson	795aa898a3	suricata: only reload rules once the ruleset file exists On a fresh install the surirulesync file.recurse creates .gitkeep before SOC has generated all-rulesets.rules. That change satisfied the surirulereload onchanges requisite, so the reload ran with no ruleset present, failed to stat the file, and reported the state (and install) as failed. Add an onlyif guard so the reload only runs when all-rulesets.rules exists. A .gitkeep-only sync now leaves the state a clean success (onlyif condition false); once SOC writes the ruleset, the reload fires normally.	2026-07-01 15:12:54 -04:00
Josh Patterson	69d77382f1	suricata: timestamp each line of reload log output Route the reload/verify output (ours plus so-common's retry/fail lines) through a synchronous timestamping pipeline so every line in reload.log is prefixed with a date/time, and preserve the real exit code via PIPESTATUS.	2026-07-01 15:12:53 -04:00
Jorge Reyes	dc9b4f3ce5	Merge pull request #16027 from Security-Onion-Solutions/reyesj2-patch-6 increase wait_for_so-kibana timeout to 10m	2026-07-01 13:48:10 -05:00
reyesj2	87b9276c79	increase wait_for_so-kibana timeout to 10m	2026-07-01 13:19:47 -05:00
Jorge Reyes	99118f9bed	Merge pull request #16023 from Security-Onion-Solutions/reyesj2/uekairgap update airgap soup to sync uek repo from iso and retain latest packag…	2026-07-01 13:14:55 -05:00
reyesj2	24b75b4a2b	typo	2026-07-01 12:50:23 -05:00
Jorge Reyes	395bd627f1	Merge pull request #16024 from Security-Onion-Solutions/reyesj2/fixsearch remove outdated eval script and associated salt utility state	2026-07-01 11:59:00 -05:00
reyesj2	c33db9d00f	remove outdated eval script and associated salt utility state	2026-07-01 11:12:39 -05:00
reyesj2	e88eb65a44	keep old packages for rollback ability	2026-07-01 10:29:05 -05:00
reyesj2	dc8c80633b	update airgap soup to sync uek repo from iso and retain latest packages only	2026-07-01 10:23:04 -05:00
Josh Patterson	f441d98e71	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-07-01 10:34:56 -04:00
Josh Patterson	895aa18486	Merge pull request #16021 from Security-Onion-Solutions/surirulereload suricata: verify reloaded ruleset is newer than the rules file	2026-07-01 10:33:14 -04:00
Josh Patterson	ee36f5f84c	suricata: verify reloaded ruleset is newer than the rules file Treating an in-progress reload as instant success could report success while Suricata was still running a stale ruleset (the in-flight reload may have started before the new all-rulesets.rules was written). Make success conditional on Suricata actually having loaded the current ruleset: capture the rules-file mtime up front, trigger a blocking reload-rules, then query ruleset-reload-time and only succeed when last_reload >= mtime. An in-progress reload now retries (waits for it to clear so our own fresh reload runs) instead of short-circuiting, and a ruleset that never catches up within the retry window fails via fail(). Also drop the redundant ruleset-reload-nonblocking call (the verified blocking reload is authoritative and the async call was what left a reload running) and log human-readable timestamps.	2026-07-01 09:00:36 -04:00
Jorge Reyes	a3f586cf88	Merge pull request #16018 from Security-Onion-Solutions/reyesj2/kf	2026-06-30 14:46:22 -05:00
Josh Patterson	52574e21c6	suricata: treat in-progress rule reload as success so-suricata-reload-rules failed the surirulereload state when a rule reload was already running: suricatasc returns {"message":"Reload already in progress","return":"NOK"}, which never matched the expected output, so retry looped all 60 attempts (~3 min) and called fail. Wrap the suricatasc calls so an in-progress reload is treated as success (the in-flight reload picks up the new rules) while genuine container-not-ready conditions still retry and ultimately fail.	2026-06-30 09:40:23 -04:00
Josh Patterson	a330bea25e	Rename push-detection beacons to clearer names Rename the two custom push-detection beacons for clarity: - pillar_db -> postgres_pillar_beacon - rules_db -> rules_beacon Salt resolves a beacon by its config-key name to a _beacons/ module of the same filename and tags its events salt/beacon/<minion>/<name>/<tag>, so each rename touches the module file, the beacon config key in beacons_pushstate.conf.jinja, and the reactor tag patterns in reactor_pushstate.conf together. Watermark filenames and log prefixes are updated to match; reactor run() logic is unchanged.	2026-06-29 14:29:07 -04:00
Josh Patterson	33c24cd136	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-06-26 15:42:56 -04:00
Josh Patterson	12f4447875	Replace inotify rule-watch beacon with poll-based rules_db beacon Salt's stock inotify beacon leaks one kernel inotify instance every time the minion rebuilds the beacon loader's __context__ (the orphaned pyinotify.Notifier is never stopped), accumulating against fs.inotify.max_user_instances=128 until inotify_init() fails with EMFILE and rule-change push detection silently stops. This is independent of disable_during_state_run. Add a custom poll-based beacon (salt/_beacons/rules_db.py) modeled on pillar_db.py: it fingerprints the suricata/strelka rule dirs each interval (relpath + mtime_ns + size, temp files excluded) against a per-dir watermark, emitting an event only on change. It holds zero inotify instances, so the leak is impossible, and it keeps firing during state runs. Swap the inotify beacon config and reactor tag mappings accordingly; the push_suricata/push_strelka reactors are unchanged (they read only data['path']).	2026-06-26 15:40:32 -04:00
Jorge Reyes	576c7bfedd	Merge pull request #16013 from Security-Onion-Solutions/reyesj2/so-start update so-stop \| so-start \| so-restart scripts	2026-06-26 13:47:09 -05:00
reyesj2	b3b7ecdded	update so-stop \| so-start \| so-restart scripts	2026-06-26 13:19:18 -05:00
Josh Patterson	da94788255	Move highstate_interval_hours to salt.schedule and split schedule.sls highstate_interval_hours describes the per-minion highstate schedule, not the active-push pipeline, so relocate it from salt.auto_apply to a new salt.schedule settings subtree. Repoint so-salt-minion-check at the new pillar path (it had been left on the stale global:push path) so its restart grace period tracks the schedule again. - Add salt.schedule.highstate_interval_hours to defaults.yaml/soc_salt.yaml and a side-effect-free salt/salt/schedule.map.jinja (SCHEDULEMERGED), matching the MERGED map convention. Consumers read SCHEDULEMERGED.highstate_interval_hours. - Split salt/schedule.sls into salt/salt/highstate_schedule.sls (every minion) and salt/salt/push_drain_schedule.sls (managers); update top.sls to apply the highstate schedule via '' and the drainer schedule via the configured-manager block. Remove the now-empty schedule.sls aggregator. - pillar_push_map.yaml and so-push-drainer: comment/doc updates only.	2026-06-26 10:51:57 -04:00
Josh Patterson	fa2ae1b87f	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-06-25 11:45:03 -04:00
Josh Patterson	5bf9751adf	do not disable during state run	2026-06-25 11:44:38 -04:00
Josh Patterson	3effdbc91e	do not disable during state run	2026-06-25 11:36:52 -04:00
Josh Patterson	8836529496	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-06-25 08:13:32 -04:00
Josh Patterson	b09c3776b7	Point pillar_db beacon at securityonion database The SOC postgres database was renamed so_soc -> securityonion (see POSTGRES_DB in salt/postgres/enabled.sls and the SOC postgres config in salt/soc/defaults.yaml). The pillar_db beacon still hardcoded so_soc, so every poll failed with 'database "so_soc" does not exist' (rc=2), silently disabling active-push detection of audit_settings changes. Update DATABASE to 'securityonion' and refresh the now-stale so_soc references in the beacon and push_pillar reactor comments.	2026-06-24 16:51:32 -04:00
Josh Patterson	dfdb1fbaeb	Move global.push config to salt.auto_apply The active-push tunables (enabled, highstate_interval_hours, debounce_seconds, drain_interval, batch, batch_wait) described how Salt auto-applies changes, not general grid config, so relocate them from the global namespace to a new salt.auto_apply settings module. - Add salt/salt/{defaults.yaml,auto_apply.map.jinja,soc_salt.yaml,adv_salt.yaml}. auto_apply.map.jinja is a dedicated, side-effect-free merge map (the existing salt/salt/map.jinja dereferences pillar.host.mainint at import time). - Remove the push blocks from salt/global/{defaults,soc_global}.yaml. - Register salt.soc_salt/salt.adv_salt in pillar/top.sls; seed the local pillar stubs for fresh installs (make_some_dirs) and upgrades (ensure_salt_local_pillar in soup, wired into up_to_3.2.0). - Repoint all consumers: GLOBALMERGED.push.* -> AUTOAPPLY.* (schedule, salt master, manager beacons, beacons_pushstate, orch.push_batch) and pillar.get('global:push...') -> 'salt:auto_apply...' (push reactors, so-push-drainer). - Add a salt: fleetwide-highstate entry to pillar_push_map.yaml so edits keep applying immediately, matching the prior global-namespace behavior.	2026-06-24 15:17:48 -04:00
Josh Patterson	61aa963a2d	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-06-24 08:10:27 -04:00
Josh Patterson	d71e80cf66	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-06-23 10:32:32 -04:00
Josh Patterson	33a116357d	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-06-10 08:56:17 -04:00
Josh Patterson	8c17ae0f66	move so-salt-minion-wait	2026-06-01 14:48:54 -04:00
Josh Patterson	f54939b444	Replace inotify pillar watch with postgres audit_settings beacon The active-push feature detected pillar/settings changes via an inotify beacon on the manager watching /opt/so/saltstack/local/pillar. Replace that pillar watch with a custom salt beacon (pillar_db) that polls the SOC so_soc.audit_settings table on a monotonic id watermark, so changes made through SOC drive immediate pushes from the database instead of the files. The suricata/strelka rule inotify watches (and pyinotify) are kept unchanged, since rule-file edits are not recorded in audit_settings. - salt/_beacons/pillar_db.py: new beacon. Polls audit_settings via `docker exec so-postgres psql` (unix-socket trust auth), tracks the last processed id in /opt/so/state/pillar_db_watch.id, seeds to MAX(id) on first run (no history replay), and emits one event per new row. - salt/reactor/push_pillar.sls: consume setting_id/node_id from the beacon event instead of a file path. App = first dotted segment of setting_id, looked up in pillar_push_map.yaml. Empty node_id -> grid-wide actions as is; populated node_id -> the app's state(s) retargeted to that one node. - salt/manager/files/beacons_pushstate.conf.jinja: drop the pillar inotify block, add the pillar_db beacon (interval = push.drain_interval); keep the suricata/strelka inotify watches. - salt/salt/files/reactor_pushstate.conf: map salt/beacon/*/pillar_db/ audit_settings to push_pillar.sls; remove the pillar inotify reactor lines; keep suricata/strelka. The intent -> so-push-drainer -> orch.push_batch pipeline is unchanged. Verified end-to-end on a standalone: a grid-wide telegraf.output change re-applied telegraf fleetwide (container replaced), and a per-host ntp.config.servers change applied ntp to only that node.	2026-05-29 14:55:13 -04:00
Josh Patterson	d48a22e37e	Merge pull request #15944 from Security-Onion-Solutions/jertel/wip Jertel/wip	2026-05-28 14:01:42 -04:00
Josh Patterson	6393d08e86	merge	2026-05-27 08:59:28 -04:00
Josh Patterson	730c828bec	Merge remote-tracking branch 'origin/jertel/wip' into saltthangs	2026-05-19 10:23:45 -04:00
Josh Patterson	b4e5171415	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-05-14 08:03:45 -04:00
Josh Patterson	84decc1db6	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-05-13 14:09:15 -04:00
Josh Patterson	7d4d6a0756	prune images if so-docker-prune exists	2026-05-08 10:13:15 -04:00
Josh Patterson	66c0a662fc	convert wait to script	2026-05-08 09:26:42 -04:00
Josh Patterson	778cc055ea	wait for salt-minion service to be ready before finishing state run	2026-05-07 17:01:20 -04:00
Josh Patterson	932deab751	update the push map	2026-05-07 10:51:53 -04:00
Josh Patterson	1281f0ee37	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-05-06 09:46:12 -04:00
Josh Patterson	f774334b6c	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-05-06 08:16:41 -04:00
Josh Patterson	7fcace34c4	add sensoroni to push map	2026-04-30 16:09:08 -04:00
Josh Patterson	9541024eb7	fix broken things	2026-04-30 15:35:24 -04:00
Josh Patterson	0d166ef732	remove trailing slashes	2026-04-30 09:53:00 -04:00
Josh Patterson	f7d2994f8b	filter temp files	2026-04-30 09:16:22 -04:00
Josh Patterson	8f0757606d	include salt..minion	2026-04-29 16:42:19 -04:00
Josh Patterson	0a8f2e01a0	install pyinotify	2026-04-29 16:41:56 -04:00
Josh Patterson	4546d7bc52	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-04-29 14:28:19 -04:00
Josh Patterson	17849d8758	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-04-28 15:49:22 -04:00
Josh Patterson	d3d30a587c	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-04-28 15:30:31 -04:00
Josh Patterson	034711d148	Merge remote-tracking branch 'origin/3/dev' into saltthangs	2026-04-28 10:47:29 -04:00
Mike Reeves	a0cf0489d6	reduce highstate frequency with active push for rules and pillars - schedule highstate every 2 hours (was 15 minutes); interval lives in global:push:highstate_interval_hours so the SOC admin UI can tune it and so-salt-minion-check derives its threshold as (interval + 1) * 3600 - add inotify beacon on the manager + master reactor + orch.push_batch that writes per-app intent files, with a so-push-drainer schedule on the manager that debounces, dedupes, and dispatches a single orchestration - pillar_push_map.yaml allowlists the apps whose pillar changes trigger an immediate targeted state.apply (targets verified against salt/top.sls); edits under pillar/minions/ trigger a state.highstate on that one minion - host-batch every push orchestration (batch: 25%, batch_wait: 15) so rule changes don't thundering-herd large fleets - new global:push:enabled kill-switch tears down the beacon, reactor config, and drainer schedule on the next highstate for operators who want to keep highstate-only behavior - set restart_policy: unless-stopped on 23 container states so docker recovers crashes without waiting for the next highstate; leave registry (always), strelka/backend (on-failure), kratos, and hydra alone with inline comments explaining why	2026-04-10 15:43:16 -04:00
Jason Ertel	613d31c8a6	merge	2026-03-05 11:52:09 -05:00