clear stale docker network endpoints before starting containers

docker_container.running fails with an address-already-in-use error when a
container holds a stale sobridge endpoint - either after its IP changes in
pillar or after an unclean shutdown leaves the endpoint behind.

Add the clear_stale_endpoint macro and call it from every enabled.sls whose
container attaches to sobridge with a static IP. The macro force-disconnects
the container from the network only when its current endpoint IP does not
match the one salt is about to assign, so it is a no-op on a normal highstate.

The strelka states use state IDs that differ from their container names, so
those calls pass state_id explicitly. nginx resolves its IP through
container_config, which varies by role.

salt/common/tools/sbin/so-restart

@@ -35,6 +35,9 @@ case $1 in
     "elastic-fleet"|"elasticfleet")
         docker_check_running "elastic-fleet" "--stop"
         docker rm "so-elastic-fleet" 2> /dev/null
+        # Removing the elastic fleet state directory, so that the next startup re-enrolls with a fresh policy
+        rm -rf /opt/so/conf/elastic-fleet/state
+
         salt-call state.apply elasticfleet queue=True
         ;;
     *)

salt/common/tools/sbin/so-stop

@@ -29,6 +29,8 @@ case $1 in
 	"elasticfleet"|"elastic-fleet")
         docker_check_running "elastic-fleet" "--stop"
         docker rm "so-elastic-fleet" 2> /dev/null
+        # Removing the elastic fleet state directory, so that the next startup re-enrolls with a fresh policy
+        rm -rf /opt/so/conf/elastic-fleet/state
         ;;
     *)
         docker_check_running "$1" "--stop"

salt/docker/macros/docker_endpoint.jinja

@@ -0,0 +1,12 @@
+{% macro clear_stale_endpoint(container, network, ipv4, state_id=None) %}
+{{ container }}_{{ network }}_stale_endpoint:
+  cmd.run:
+    - name: docker network disconnect -f {{ network }} {{ container }} || true
+    - onlyif: docker inspect {{ container }}
+    - unless: >-
+        docker inspect -f
+        '{{ '{{' }} with index .NetworkSettings.Networks "{{ network }}" {{ '}}' }}{{ '{{' }} .IPAMConfig.IPv4Address {{ '}}' }}{{ '{{' }} end {{ '}}' }}'
+        {{ container }} 2>/dev/null | grep -qx '{{ ipv4 }}'
+    - require_in:
+      - docker_container: {{ state_id or container }}
+{% endmacro %}

salt/elastalert/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-elastalert', 'sobridge', DOCKERMERGED.containers['so-elastalert'].ip) }}
 
 include:
   - elastalert.config

salt/elastic-fleet-package-registry/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-elastic-fleet-package-registry', 'sobridge', DOCKERMERGED.containers['so-elastic-fleet-package-registry'].ip) }}
 
 include:
   - elastic-fleet-package-registry.config

salt/elasticagent/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-elastic-agent', 'sobridge', DOCKERMERGED.containers['so-elastic-agent'].ip) }}
 
 include:
   - ca

salt/elasticfleet/enabled.sls

@@ -8,9 +8,14 @@
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'elasticfleet/map.jinja' import ELASTICFLEETMERGED %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
 
 {#   This value is generated during node install and stored in minion pillar #}
 {%   set SERVICETOKEN = salt['pillar.get']('elasticfleet:config:server:es_token','') %}
+{#   Prevent Elastic Agent from re-enrolling with a new agent.id everytime the container starts up.
+       - if a fresh enrollment is needed use 'so-stop elasticfleet'
+       #}
+{%   set ENROLLED = salt['file.file_exists']('/opt/so/conf/elastic-fleet/state/fleet.enc') %}
 
 include:
   - ca
@@ -39,6 +44,8 @@ elasticagent_syncartifacts:
 {% endif %}
 
 {%   if SERVICETOKEN != '' %}
+{{ clear_stale_endpoint('so-elastic-fleet', 'sobridge', DOCKERMERGED.containers['so-elastic-fleet'].ip) }}
+
 so-elastic-fleet:
   docker_container.running:
     - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-elastic-agent:{{ GLOBALS.so_version }}
@@ -65,6 +72,7 @@ so-elastic-fleet:
       - /etc/pki/elasticfleet-server.crt:/etc/pki/elasticfleet-server.crt:ro
       - /etc/pki/elasticfleet-server.key:/etc/pki/elasticfleet-server.key:ro
       - /etc/pki/tls/certs/intca.crt:/etc/pki/tls/certs/intca.crt:ro
+      - /opt/so/conf/elastic-fleet/state:/usr/share/elastic-agent/state
       - /opt/so/log/elasticfleet:/usr/share/elastic-agent/logs 
      {% if DOCKERMERGED.containers['so-elastic-fleet'].custom_bind_mounts %}
         {% for BIND in DOCKERMERGED.containers['so-elastic-fleet'].custom_bind_mounts %}
@@ -72,6 +80,7 @@ so-elastic-fleet:
         {% endfor %}
       {% endif %}      
     - environment:
+      {% if not ENROLLED %}
       - FLEET_SERVER_ENABLE=true
       - FLEET_URL=https://{{ GLOBALS.hostname }}:8220
       - FLEET_SERVER_ELASTICSEARCH_HOST=https://{{ GLOBALS.manager }}:9200
@@ -81,6 +90,9 @@ so-elastic-fleet:
       - FLEET_SERVER_CERT_KEY=/etc/pki/elasticfleet-server.key
       - FLEET_CA=/etc/pki/tls/certs/intca.crt     
       - FLEET_SERVER_ELASTICSEARCH_CA=/etc/pki/tls/certs/intca.crt
+      {% endif %}
+      - STATE_PATH=/usr/share/elastic-agent/state
+      - CONFIG_PATH=/usr/share/elastic-agent/state
       - LOGS_PATH=logs
       {% if DOCKERMERGED.containers['so-elastic-fleet'].extra_env %}
         {% for XTRAENV in DOCKERMERGED.containers['so-elastic-fleet'].extra_env %}
@@ -99,6 +111,7 @@ so-elastic-fleet:
       - x509: etc_elasticfleet_crt
     - require:
       - file: trusttheca
+      - file: eastatedir
       - x509: etc_elasticfleet_key
       - x509: etc_elasticfleet_crt
 

salt/elasticfleet/install_agent_grid.sls

@@ -10,6 +10,15 @@
 {% set AGENT_STATUS = salt['service.available']('elastic-agent') %}
 {% set AGENT_EXISTS = salt['file.file_exists']('/opt/Elastic/Agent/elastic-agent') %}
 
+so-elastic-agent-install:
+  file.managed:
+    - name: /usr/sbin/so-elastic-agent-install
+    - source: salt://elasticfleet/tools/sbin/so-elastic-agent-install
+    - user: 947
+    - group: 939
+    - mode: 755
+    - show_changes: False
+
 {% if not AGENT_STATUS or not AGENT_EXISTS %}
 
 pull_agent_installer:
@@ -21,11 +30,9 @@ pull_agent_installer:
 
 run_installer:
   cmd.run:
-    - name: ./so-elastic-agent_linux_amd64 -token={{ GRIDNODETOKEN }} -force
-    - cwd: /opt/so
-    - retry:
-        attempts: 3
-        interval: 20
+    - name: /usr/sbin/so-elastic-agent-install "{{ GRIDNODETOKEN }}"
+    - require:
+      - file: pull_agent_installer
 
 cleanup_agent_installer:
   file.absent:

salt/elasticfleet/tools/sbin/so-elastic-agent-install

@@ -0,0 +1,100 @@
+#!/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.
+
+. /usr/sbin/so-elastic-fleet-common
+
+
+# passed in as arg from elasticfleet/install_agent_grid.sls, else pulled from pillar later
+GRIDNODETOKEN="$1"
+LOGFILE="/opt/so/SO-Elastic-Agent_Installer_Health.log"
+
+check_agent_health() {
+    timeout=300
+    interval=10
+    start=$SECONDS
+
+    while (( SECONDS - start < timeout )); do
+        agent_status=$(elastic-agent status 2>&1)
+        echo -e "\n$(date)\n$agent_status\n" >> "$LOGFILE"
+        if echo "$agent_status" | grep -A1 'elastic-agent$' | grep -q 'status: (HEALTHY)'; then
+            return 0
+        fi
+        echo "The Elastic Agent is not yet healthy. Waiting for ${interval} seconds before checking again..."
+        sleep "$interval"
+    done
+
+    echo "The Elastic Agent did not become healthy within ${timeout} seconds"
+    return 1
+}
+
+uninstall_agent() {
+    if command -v elastic-agent >/dev/null 2>&1; then
+        elastic-agent uninstall -f
+    fi
+}
+
+if [[ -z "$GRIDNODETOKEN" ]]; then
+    noderole=$(so-yaml.py get -r /etc/salt/grains role)
+    if [[ "$noderole" == "so-heavynode" ]]; then
+        GRIDNODETOKEN=$(salt-call pillar.get global:fleet_grid_enrollment_token_heavy --out=newline_values_only)
+    else
+        GRIDNODETOKEN=$(salt-call pillar.get global:fleet_grid_enrollment_token_general --out=newline_values_only)
+    fi
+fi
+
+if [[ -z "$GRIDNODETOKEN" ]]; then
+    echo "Unable to determine Elastic Fleet enrollment token. Exiting."
+
+    exit 1
+fi
+
+if [[ ! -x /opt/so/so-elastic-agent_linux_amd64 ]]; then
+    echo "Downloading so-elastic-agent installer... This could take a while if another Salt job is running."
+
+    # When running outside of elasticfleet/install_agent_grid.sls we need to download the installer independently.
+    #  PYTHONWARNINGS="ignore" to avoid messages like the following when running salt-call:
+    #  '/opt/saltstack/salt/lib/python3.10/site-packages/salt/transport/base.py:129: TransportWarning: Unclosed transport! <salt.transport.zeromq.RequestClient object at 0x7fc5f0ee7a30>
+    #    File "/bin/salt-call", line 12, in <module>
+    #      sys.exit(salt_call())'
+
+    PYTHONWARNINGS="ignore" salt-call state.single file.managed name=/opt/so/so-elastic-agent_linux_amd64 source=salt://elasticfleet/files/so_agent-installers/so-elastic-agent_linux_amd64 mode=755 makedirs=True queue=True
+
+fi
+
+if [[ -x /opt/so/so-elastic-agent_linux_amd64 ]]; then
+    attempts=0
+    cd /opt/so/ || exit 1
+
+    truncate -s 0 "$LOGFILE"
+
+    uninstall_agent
+
+    while [[ $attempts -lt 3 ]]; do
+        if ./so-elastic-agent_linux_amd64 -token="$GRIDNODETOKEN" -force && echo "Verifying Elastic Agent health..." && check_agent_health; then
+            rm -f /opt/so/so-elastic-agent_linux_amd64
+            elastic-agent status
+
+            exit 0
+        fi
+
+        attempts=$((attempts + 1))
+
+        if [[ $attempts -lt 3 ]]; then
+            echo "Unable to verify Elastic Agent health... Retrying in 20 seconds..."
+            sleep 20
+        fi
+    done
+
+    uninstall_agent
+    rm -f /opt/so/so-elastic-agent_linux_amd64
+    echo "The so-elastic-agent installer failed after 3 attempts. Exiting."
+
+    exit 1
+else
+    echo "Unable to locate so-elastic-agent installer. Exiting."
+
+    exit 1
+fi

salt/elasticfleet/tools/sbin_jinja/so-elastic-agent-gen-installers

@@ -30,7 +30,7 @@ done
 if [[ -z $FLEETHOST ]] || [[ -z $ENROLLMENTOKEN ]]; then 
   printf "\nFleet Host URL, Enrollment Token or Elastic Version empty - exiting..."
   printf "\nFleet Host: $FLEETHOST, Enrollment Token: $ENROLLMENTOKEN\n"
-  exit 
+  exit 1
 fi
 
 OSARCH=( "linux-x86_64" "windows-x86_64" "darwin-x86_64" "darwin-aarch64" )
@@ -62,31 +62,54 @@ do
 done 
 
 GOTARGETOS=( "linux" "windows" "darwin" "darwin/arm64" )
-GOARCH="amd64"
 printf "\n### Generating OS packages using the cleaned up tarballs"
-for GOOS in "${GOTARGETOS[@]}"
-do
+for GOOS in "${GOTARGETOS[@]}"; do
+    GOARCH="amd64"
     if [[ $GOOS == 'darwin/arm64' ]]; then GOOS="darwin" && GOARCH="arm64"; fi
     printf "\n\n### Generating $GOOS/$GOARCH Installer...\n"
     docker run -e CGO_ENABLED=0 -e GOOS=$GOOS -e GOARCH=$GOARCH \
     --mount type=bind,source=/etc/pki/tls/certs/,target=/workspace/files/cert/ \
     --mount type=bind,source=/nsm/elastic-agent-workspace/,target=/workspace/files/elastic-agent/ \
-    --mount type=bind,source=/opt/so/saltstack/local/salt/elasticfleet/files/so_agent-installers/,target=/output/ \
+    --mount type=bind,source=/opt/so/saltstack/local/salt/elasticfleet/files/,target=/output/ \
     {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-elastic-agent-builder:{{ GLOBALS.so_version }} go build -ldflags "-X main.fleetHostURLsList=$FLEETHOST -X main.enrollmentToken=$ENROLLMENTOKEN"  -o /output/so-elastic-agent_${GOOS}_${GOARCH}
     printf "\n### $GOOS/$GOARCH Installer Generated...\n"
 done
 
 printf "\n\n### Generating MSI...\n"
-cp /opt/so/saltstack/local/salt/elasticfleet/files/so_agent-installers/so-elastic-agent_windows_amd64 /opt/so/saltstack/local/salt/elasticfleet/files/so_agent-installers/so-elastic-agent_windows_amd64.exe
+cp /opt/so/saltstack/local/salt/elasticfleet/files/so-elastic-agent_windows_amd64 /opt/so/saltstack/local/salt/elasticfleet/files/so-elastic-agent_windows_amd64.exe
 docker run \
---mount type=bind,source=/opt/so/saltstack/local/salt/elasticfleet/files/so_agent-installers/,target=/output/ -w /output \
+--mount type=bind,source=/opt/so/saltstack/local/salt/elasticfleet/files/,target=/output/ -w /output \
 {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-elastic-agent-builder:{{ GLOBALS.so_version }} wixl -o so-elastic-agent_windows_amd64_msi --arch x64 /workspace/so-elastic-agent.wxs
 printf "\n### MSI Generated...\n"
 
+# Verify installers were created
+for GOOS in "${GOTARGETOS[@]}"; do
+    GOARCH="amd64"
+    if [[ $GOOS == 'darwin/arm64' ]]; then GOOS="darwin"; GOARCH="arm64"; fi
+    if [[ ! -f /opt/so/saltstack/local/salt/elasticfleet/files/so-elastic-agent_${GOOS}_${GOARCH} ]]; then
+        printf "\n### ERROR: Installer for %s/%s was not generated. Exiting...\n" "$GOOS" "$GOARCH"
+        exit 1
+    fi
+    # After verifying new installer was generated, move it to so_agent-installers directory
+    mv /opt/so/saltstack/local/salt/elasticfleet/files/so-elastic-agent_${GOOS}_${GOARCH} /opt/so/saltstack/local/salt/elasticfleet/files/so_agent-installers/
+done
+
+# Verify MSI installer
+if [[ ! -f /opt/so/saltstack/local/salt/elasticfleet/files/so-elastic-agent_windows_amd64_msi ]]; then
+    printf "\n### ERROR: Installer MSI was not generated. Exiting...\n"
+    exit 1
+else
+    # After verifying new installer MSI was generated, move it to so_agent-installers directory
+    mv /opt/so/saltstack/local/salt/elasticfleet/files/so-elastic-agent_windows_amd64_msi /opt/so/saltstack/local/salt/elasticfleet/files/so_agent-installers/
+fi
+
 printf "\n### Cleaning up temp files \n"
 rm -rf /nsm/elastic-agent-workspace
-rm -rf /opt/so/saltstack/local/salt/elasticfleet/files/so_agent-installers/so-elastic-agent_windows_amd64.exe
+rm -rf /opt/so/saltstack/local/salt/elasticfleet/files/so-elastic-agent_windows_amd64.exe
 
 printf "\n### Copying so_agent-installers to /nsm/elastic-fleet/ for nginx.\n"
 \cp -vr /opt/so/saltstack/local/salt/elasticfleet/files/so_agent-installers/ /nsm/elastic-fleet/
 chmod 644 /nsm/elastic-fleet/so_agent-installers/*
+
+# if we got here all installers have been generated successfully
+exit 0

salt/elasticfleet/tools/sbin_jinja/so-elastic-fleet-setup

@@ -244,11 +244,37 @@ printf '%s\n'\
     "" >> "$global_pillar_file"
 
 # Call Elastic-Fleet Salt State
-printf "\nApplying elasticfleet state"
-salt-call state.apply elasticfleet queue=True
+printf "\nApplying elasticfleet state\n"
+for state_attempt in {1..3}; do
+    if salt-call state.apply elasticfleet queue=True; then
+        break
+    elif [[ $state_attempt -lt 3 ]]; then
+        printf "\nElasticfleet state did not complete successfully... Attempt (%s/3). Retrying...\n" "$state_attempt"
+        sleep 10
+    else
+        printf "\nFailure(s) in elasticfleet state... Exiting...\n"
+        exit 1
+    fi
+done
 
+printf "\nRunning so-elastic-agent-gen-installers\n"
 # Generate installers & install Elastic Agent on the node
-so-elastic-agent-gen-installers
-printf "\nApplying elasticfleet.install_agent_grid state"
-salt-call state.apply elasticfleet.install_agent_grid queue=True
-exit 0
+for agent_gen_attempt in {1..3}; do
+    if so-elastic-agent-gen-installers; then
+        break
+    elif [[ $agent_gen_attempt -lt 3 ]]; then
+        printf "\nUnable to generate Elastic Agent installers... Attempt (%s/3). Retrying...\n" "$agent_gen_attempt"
+        sleep 10
+    else
+        printf "\nFailed to generate Elastic Agent installers after 3 attempts. Exiting...\n"
+        exit 1
+    fi
+done
+
+printf "\nApplying elasticfleet.install_agent_grid state\n"
+if ! salt-call state.apply elasticfleet.install_agent_grid queue=True; then
+    printf "\nFailure(s) in elasticfleet.install_agent_grid state... Exiting...\n"
+    exit 1
+fi
+
+printf "\nElastic Fleet setup completed successfully\n"

salt/elasticsearch/enabled.sls

@@ -10,6 +10,9 @@
 {%   from 'elasticsearch/config.map.jinja' import ELASTICSEARCH_NODES %}
 {%   from 'elasticsearch/config.map.jinja' import ELASTICSEARCH_SEED_HOSTS %}
 {%   from 'elasticsearch/config.map.jinja' import ELASTICSEARCHMERGED %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-elasticsearch', 'sobridge', DOCKERMERGED.containers['so-elasticsearch'].ip) }}
 
 include:
   - ca

salt/elasticsearch/soc_elasticsearch.yaml

@@ -28,14 +28,14 @@ elasticsearch:
       description: The maximum number of memory map areas a process may use. Elasticsearch uses a mmapfs directory by default to store its indices. The default operating system limits on mmap counts could be too low, which may result in out of memory exceptions.
       forcedType: int
       helpLink: elasticsearch
-  retention: 
+  retention:
     retention_pct:
       decription: Total percentage of space used by Elasticsearch for multi node clusters
       helpLink: elasticsearch
       global: True
   config:
     cluster:
-      name: 
+      name:
         description: The name of the Security Onion Elasticsearch cluster, for identification purposes.
         readonly: True
         global: True
@@ -55,13 +55,13 @@ elasticsearch:
               forcedType: bool
               helpLink: elasticsearch
             watermark:
-              low: 
+              low:
                 description: The lower percentage of used disk space representing a healthy node.
                 helpLink: elasticsearch
-              high: 
+              high:
                 description: The higher percentage of used disk space representing an unhealthy node.
                 helpLink: elasticsearch
-              flood_stage: 
+              flood_stage:
                 description: The max percentage of used disk space that will cause the node to take protective actions, such as blocking incoming events.
                 helpLink: elasticsearch
     action:
@@ -172,11 +172,11 @@ elasticsearch:
                 forcedType: int
                 global: True
                 helpLink: elasticsearch
-              refresh_interval: 
+              refresh_interval:
                 description: Seconds between index refreshes. Shorter intervals can cause query performance to suffer since this is a synchronous and resource-intensive operation.
                 global: True
                 helpLink: elasticsearch
-              number_of_shards: 
+              number_of_shards:
                 description: Number of shards required for this index. Using multiple shards increases fault tolerance, but also increases storage and network costs.
                 global: True
                 helpLink: elasticsearch
@@ -269,7 +269,7 @@ elasticsearch:
                   global: True
                   advanced: True
           warm:
-            min_age: 
+            min_age:
               description: Minimum age of index. ex. 30d - This determines when the index should be moved to the warm tier. Nodes in the warm tier generally don’t need to be as fast as those in the hot tier. It’s important to note that this is calculated relative to the rollover date (NOT the original creation date of the index). For example, if you have an index that is set to rollover after 30 days and warm min_age set to 30 then there will be 30 days from index creation to rollover and then an additional 30 days before moving to warm tier.
               regex: ^[0-9]{1,5}d$
               forcedType: string
@@ -370,7 +370,7 @@ elasticsearch:
         template:
           settings:
             index:
-              number_of_replicas: 
+              number_of_replicas:
                 description: Number of replicas required for this index. Multiple replicas protects against data loss, but also increases storage costs.
                 forcedType: int
                 global: True
@@ -391,12 +391,12 @@ elasticsearch:
                     global: True
                     advanced: True
                     helpLink: elasticsearch
-              refresh_interval: 
+              refresh_interval:
                 description: Seconds between index refreshes. Shorter intervals can cause query performance to suffer since this is a synchronous and resource-intensive operation.
                 global: True
                 advanced: True
                 helpLink: elasticsearch
-              number_of_shards: 
+              number_of_shards:
                 description: Number of shards required for this index. Using multiple shards increases fault tolerance, but also increases storage and network costs.
                 global: True
                 advanced: True
@@ -645,6 +645,7 @@ elasticsearch:
             global: True
             advanced: True
             helpLink: elasticsearch
+    so-logs-soc: *dataStreamSettings
     so-logs-system_x_auth: *dataStreamSettings
     so-logs-system_x_syslog: *dataStreamSettings
     so-logs-system_x_system: *dataStreamSettings

salt/hydra/enabled.sls

@@ -14,6 +14,9 @@
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
 {%   if 'api' in salt['pillar.get']('features', []) %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-hydra', 'sobridge', DOCKERMERGED.containers['so-hydra'].ip) }}
 
 include:
   - hydra.config

salt/influxdb/enabled.sls

@@ -9,6 +9,9 @@
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   set PASSWORD = salt['pillar.get']('secrets:influx_pass') %}
 {%   set TOKEN = salt['pillar.get']('influxdb:token') %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-influxdb', 'sobridge', DOCKERMERGED.containers['so-influxdb'].ip) }}
 
 include:
   - influxdb.ssl

salt/kafka/enabled.sls

@@ -16,6 +16,9 @@
 {%   set KAFKANODES = salt['pillar.get']('kafka:nodes') %}
 {%   set KAFKA_EXTERNAL_ACCESS = salt['pillar.get']('kafka:config:external_access:enabled', default=False) %}
 {%   if 'gmd' in salt['pillar.get']('features', []) %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-kafka', 'sobridge', DOCKERMERGED.containers['so-kafka'].ip) }}
 
 include:
   - kafka.ca

salt/kibana/enabled.sls

@@ -8,6 +8,9 @@
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'elasticsearch/config.map.jinja' import ELASTICSEARCHMERGED %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-kibana', 'sobridge', DOCKERMERGED.containers['so-kibana'].ip) }}
 
 include:
   - kibana.config

salt/kratos/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-kratos', 'sobridge', DOCKERMERGED.containers['so-kratos'].ip) }}
 
 include:
   - kratos.config

salt/logstash/enabled.sls

@@ -10,6 +10,9 @@
 {%   from 'logstash/map.jinja' import LOGSTASH_MERGED %}
 {%   from 'logstash/map.jinja' import LOGSTASH_NODES %}
 {%   set lsheap = LOGSTASH_MERGED.settings.lsheap %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-logstash', 'sobridge', DOCKERMERGED.containers['so-logstash'].ip) }}
 
 include:
   - ca

salt/manager/tools/sbin/soup

@@ -784,12 +784,23 @@ post_to_3.1.0() {
 
 ### 3.2.0 Scripts ###
 
+recollate_postgres() {
+  echo ""
+  echo "Recollating PostgreSQL databases. The following output may contain warnings about a version mismatch, followed by a note indicating that the collation version has been changed."
+  for db in postgres securityonion so_telegraf; do
+    docker exec so-postgres psql -U postgres $db -c "reindex database $db"
+    docker exec so-postgres psql -U postgres $db -c "alter database $db refresh collation version"
+  done
+  echo "Recollating PostgreSQL databases complete."
+  echo ""
+}
+
 bootstrap_so_soc_database() {
   # init-db.sh is mounted into so-postgres at /docker-entrypoint-initdb.d/init-db.sh
   # and runs automatically only on a fresh data directory. Hosts upgrading from
   # 3.1.0 already have /nsm/postgres populated, so the so_soc bootstrap block
   # added in 3.2 never fires. Re-run the script explicitly; it's idempotent.
-  echo "Bootstrapping so_soc database via init-db.sh."
+  echo "Bootstrapping database via init-db.sh."
   # The postgres image has no USER directive, so `docker exec` defaults to
   # root, and the container env intentionally omits POSTGRES_USER (the upstream
   # entrypoint defaults it transiently during first-init only). Recreate both
@@ -800,10 +811,13 @@ bootstrap_so_soc_database() {
     return 0
   fi
   if ! $exec_cmd; then
-    FINAL_MESSAGE_QUEUE+=("WARNING: init-db.sh failed inside so-postgres during the 3.2.0 upgrade; the so_soc database may not have been bootstrapped. Re-run manually: $exec_cmd")
+    FINAL_MESSAGE_QUEUE+=("WARNING: init-db.sh failed inside so-postgres during the 3.2.0 upgrade; the database may not have been bootstrapped. Re-run manually: $exec_cmd")
     return 0
   fi
-  echo "so_soc bootstrap complete."
+  echo "Database bootstrap complete."
+
+  echo "Restarting so-soc container to pick up database changes"
+  docker restart so-soc
 }
 
 # Existing grids should keep ILM unless an admin explicitly opts in to DLM.
@@ -882,6 +896,9 @@ up_to_3.2.0() {
 }
 
 post_to_3.2.0() {
+  # Recollate due to image OS rebase
+  recollate_postgres
+
   bootstrap_so_soc_database
 
   # Including agent regen script here since it was missed in post_to_3.1.0

salt/nginx/enabled.sls

@@ -8,6 +8,7 @@
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'nginx/map.jinja' import NGINXMERGED %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
 
 include:
   - nginx.ssl
@@ -31,6 +32,8 @@ make-rule-dir-nginx:
 {%     set container_config = 'so-nginx-fleet-node' %}
 {%   endif %}
 
+{{ clear_stale_endpoint('so-nginx', 'sobridge', DOCKERMERGED.containers[container_config].ip) }}
+
 so-nginx:
   docker_container.running:
     - image: {{ GLOBALS.registry_host }}:5000/{{ GLOBALS.image_repo }}/so-nginx:{{ GLOBALS.so_version }}

salt/postgres/enabled.sls

@@ -8,6 +8,9 @@
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   set SO_POSTGRES_USER = salt['pillar.get']('postgres:auth:users:so_postgres_user:user', 'so_postgres') %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-postgres', 'sobridge', DOCKERMERGED.containers['so-postgres'].ip) }}
 
 include:
   - postgres.auth

salt/redis/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-redis', 'sobridge', DOCKERMERGED.containers['so-redis'].ip) }}
 
 include:
   - ca

salt/registry/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-dockerregistry', 'sobridge', DOCKERMERGED.containers['so-dockerregistry'].ip) }}
 
 include:
   - registry.ssl

salt/soc/defaults.map.jinja

@@ -8,6 +8,7 @@
 {% from 'docker/docker.map.jinja' import DOCKERMERGED -%}
 {% set INFLUXDB_TOKEN = salt['pillar.get']('influxdb:token') %}
 {% import_text 'influxdb/metrics_link.txt' as METRICS_LINK %}
+{% from 'telegraf/map.jinja' import TELEGRAFMERGED %}
 
 {% for module, application_url in GLOBALS.application_urls.items() %}
 {%   do SOCDEFAULTS.soc.config.server.modules[module].update({'hostUrl': application_url}) %}
@@ -24,13 +25,22 @@
 
 {% do SOCDEFAULTS.soc.config.server.modules.elastic.update({'username': GLOBALS.elasticsearch.auth.users.so_elastic_user.user, 'password': GLOBALS.elasticsearch.auth.users.so_elastic_user.pass}) %}
 
-{% do SOCDEFAULTS.soc.config.server.modules.influxdb.update({'hostUrl': 'https://' ~ GLOBALS.influxdb_host ~ ':8086'}) %}
-{% do SOCDEFAULTS.soc.config.server.modules.influxdb.update({'token': INFLUXDB_TOKEN}) %}
-{% for tool in SOCDEFAULTS.soc.config.server.client.tools %}
-{%   if tool.name == "toolInfluxDb" and METRICS_LINK | length > 0 %}
-{%     do tool.update({'link': METRICS_LINK}) %}
-{%   endif %}
-{% endfor %}
+{% if TELEGRAFMERGED.output == 'POSTGRES' %}
+{%   for tool in SOCDEFAULTS.soc.config.server.client.tools %}
+{%     if tool.name == "toolInfluxDb" %}
+{%       do SOCDEFAULTS.soc.config.server.client.tools.remove(tool) %}
+{%     endif %}
+{%   endfor %}
+
+{% else %}
+{%   do SOCDEFAULTS.soc.config.server.modules.influxdb.update({'hostUrl': 'https://' ~ GLOBALS.influxdb_host ~ ':8086'}) %}
+{%   do SOCDEFAULTS.soc.config.server.modules.influxdb.update({'token': INFLUXDB_TOKEN}) %}
+{%   for tool in SOCDEFAULTS.soc.config.server.client.tools %}
+{%     if tool.name == "toolInfluxDb" and METRICS_LINK | length > 0 %}
+{%       do tool.update({'link': METRICS_LINK}) %}
+{%     endif %}
+{%   endfor %}
+{% endif %}
 
 {% do SOCDEFAULTS.soc.config.server.modules.statickeyauth.update({'anonymousCidr': DOCKERMERGED.range, 'apiKey': pillar.sensoroni.config.sensoronikey}) %}
 

salt/soc/defaults.yaml

@@ -1500,18 +1500,23 @@ soc:
           playbookRepos:
             default:
               - repo: https://github.com/Security-Onion-Solutions/securityonion-resources-playbooks
+                rulesetName: sos-playbook-resources
                 branch: main
                 folder: securityonion-normalized
               - repo: https://github.com/Security-Onion-Solutions/securityonion-resources-playbooks
+                rulesetName: sos-published
                 branch: published
                 folder: sigma
             airgap:
               - repo: file:///nsm/airgap-resources/playbooks/securityonion-resources-playbooks
+                rulesetName: sos-resources-ag
                 branch: main
                 folder: securityonion-normalized
         assistant:
           systemPromptAddendum: ""
           systemPromptAddendumMaxLength: 50000
+          maxSubSessionTokens: 0
+          maxDelegationDepth: 5
           adapters:
             - name: SOAI
               protocol: securityonion_ai_cloud
@@ -1523,6 +1528,10 @@ soc:
               serviceAccountJSON: ""
               serviceAccountLocation: ""
               healthTimeoutSeconds: 5
+          agentic: false
+          agentMapping:
+            Orchestrator: sonnet
+            Hunter: sonnet
         onionconfig:
           saltstackDir: /opt/so/saltstack
           bypassEnabled: false
@@ -2692,6 +2701,8 @@ soc:
           thresholdColorRatioLow: 0.5
           thresholdColorRatioMed: 0.75
           thresholdColorRatioMax: 1
+          toolBusyMaxRetries: 30
+          toolBusyRetryDelayMs: 1000
           availableModels:
             - id: sonnet
               displayName: Claude Sonnet

salt/soc/enabled.sls

@@ -10,6 +10,9 @@
 {%   from 'soc/merged.map.jinja' import DOCKER_EXTRA_HOSTS %}
 {%   from 'soc/merged.map.jinja' import SOCMERGED %}
 
+{% from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+{{ clear_stale_endpoint('so-soc', 'sobridge', DOCKERMERGED.containers['so-soc'].ip) }}
+
 include:
   - ca
   - soc.config

salt/soc/merged.map.jinja

@@ -7,6 +7,11 @@
 {% from 'soc/defaults.map.jinja' import SOCDEFAULTS with context %}
 {% from 'elasticsearch/config.map.jinja' import ELASTICSEARCH_NODES %}
 {% from 'manager/map.jinja' import MANAGERMERGED %}
+{% from 'telegraf/map.jinja' import TELEGRAFMERGED %}
+{%- set PG_ENTRY = salt['pillar.get']('telegraf:postgres_creds:' ~ grains.id, {}) %}
+{%- set PG_USER = PG_ENTRY.get('user', '') %}
+{%- set PG_PASS = PG_ENTRY.get('pass', '') %}
+
 {% set DOCKER_EXTRA_HOSTS = ELASTICSEARCH_NODES %}
 {% do DOCKER_EXTRA_HOSTS.append({GLOBALS.influxdb_host:pillar.node_data[GLOBALS.influxdb_host].ip}) %}
 
@@ -75,6 +80,20 @@
 {%   do SOCMERGED.config.server.update({'airgapEnabled': false}) %}
 {% endif %}
 
+{# Define the postgresmetrics module if telegraf is setup to only use Postgres #}
+{% if TELEGRAFMERGED.output != 'INFLUXDB' and PG_USER and PG_PASS %}
+{%   do SOCMERGED.config.server.modules.update({
+        'postgresmetrics': {
+          'database': 'so_telegraf',
+          'host': GLOBALS.manager_ip,
+          'password': PG_PASS,
+          'port': 5432,
+          'sslMode': 'allow',
+          'user': PG_USER,
+        }
+      }) %}
+{%   do SOCMERGED.config.server.modules.pop('influxdb') %}
+{% endif %}
 
 {# Define the Detections custom ruleset that should always be present #}
 {% set CUSTOM_RULESET = {

salt/soc/soc_soc.yaml

@@ -727,6 +727,16 @@ soc:
             description: Maximum length of the system prompt addendum. Longer prompts will be truncated.
             global: True
             advanced: True
+          maxSubSessionTokens:
+            description: Maximum number of output tokens a delegated sub-session may generate across all of its turns. When the budget is reached, the sub-agent is halted and its result is returned to the parent agent. Set to 0 to disable the limit.
+            global: True
+            advanced: True
+            forcedType: int
+          maxDelegationDepth:
+            description: Maximum delegation nesting depth for sub-agents. For example, a value of 2 lets the main agent delegate to a sub-agent that may itself delegate one level deeper. Any deeper delegation is refused and the requesting agent continues without it. Set to 0 to disable the limit.
+            global: True
+            advanced: True
+            forcedType: int
           adapters:
             description: Configuration for AI adapters used by the Onion AI assistant. Please see documentation for help on which fields are required for which protocols.
             global: True
@@ -765,12 +775,29 @@ soc:
               label: Health Timeout Seconds
               required: False
               forcedType: int
+          agentic:
+            description: Indicates if the Assistant Module should operate in agentic mode or not. If true, agents can work together to solve tasks.
+            global: True
+            forcedType: bool
+          agentMapping:
+            Orchestrator:
+              description: The initial agent in most agentic conversations. This agent will delegate requests to specialized agents.
+              global: True
+            Hunter:
+              description: This agent is specialized in querying events.
+              global: True
       client:
         assistant:
           enabled:
             description: Set to true to enable the Onion AI assistant in SOC.
             global: True
             forcedType: bool
+          toolBusyMaxRetries:
+            description: How many times to retry auto approving a tool while a tool is already running.
+            global: True
+          toolBusyRetryDelayMs:
+            description: How long in milliseconds to wait between each retry when auto approving a tool.
+            global: True
           investigationPrompt: 
             description: Prompt given to Onion AI when beginning an investigation.
             global: True

salt/strelka/backend/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-strelka-backend', 'sobridge', DOCKERMERGED.containers['so-strelka-backend'].ip, state_id='strelka_backend') }}
 
 include:
   - strelka.backend.config

salt/strelka/coordinator/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-strelka-coordinator', 'sobridge', DOCKERMERGED.containers['so-strelka-coordinator'].ip, state_id='strelka_coordinator') }}
 
 include:
   - strelka.coordinator.config

salt/strelka/filestream/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-strelka-filestream', 'sobridge', DOCKERMERGED.containers['so-strelka-filestream'].ip, state_id='strelka_filestream') }}
 
 include:
   - strelka.filestream.config

salt/strelka/frontend/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-strelka-frontend', 'sobridge', DOCKERMERGED.containers['so-strelka-frontend'].ip, state_id='strelka_frontend') }}
 
 include:
   - strelka.frontend.config

salt/strelka/gatekeeper/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-strelka-gatekeeper', 'sobridge', DOCKERMERGED.containers['so-strelka-gatekeeper'].ip, state_id='strelka_gatekeeper') }}
 
 include:
   - strelka.gatekeeper.config

salt/strelka/manager/enabled.sls

@@ -7,6 +7,9 @@
 {% if sls.split('.')[0] in allowed_states %}
 {%   from 'docker/docker.map.jinja' import DOCKERMERGED %}
 {%   from 'vars/globals.map.jinja' import GLOBALS %}
+{%   from 'docker/macros/docker_endpoint.jinja' import clear_stale_endpoint %}
+
+{{ clear_stale_endpoint('so-strelka-manager', 'sobridge', DOCKERMERGED.containers['so-strelka-manager'].ip, state_id='strelka_manager') }}
 
 include:
   - strelka.manager.config

salt/telegraf/soc_telegraf.yaml

@@ -5,7 +5,7 @@ telegraf:
     advanced: True
     helpLink: influxdb
   output:
-    description: Selects the backend(s) Telegraf writes metrics to. INFLUXDB keeps the current behavior; POSTGRES writes to the grid's Postgres instance; BOTH dual-writes for migration validation.
+    description: Selects the backend(s) Telegraf writes metrics to. INFLUXDB keeps the current behavior; POSTGRES writes to the grid's Postgres instance; BOTH dual-writes for migration validation. When set to BOTH, the grid screen's metrics are pulled from Postgres, and the InfluxDB tool link remains visible. When set to POSTGRES, the InfluxDB tool link is removed.
     options:
       - INFLUXDB
       - POSTGRES

setup/so-setup

@@ -793,7 +793,7 @@ if ! [[ -f $install_opt_file ]]; then
 		logCmd "so-soc-restart"
 		title "Setting up Elastic Fleet"
 		logCmd "salt-call state.apply elasticfleet.config"	
-		if ! logCmd so-elastic-fleet-setup; then
+		if ! so-elastic-fleet-setup; then
 			fail_setup "Failed to run so-elastic-fleet-setup"
 		fi
 		mark_setup_complete