Merge pull request #15947 from Security-Onion-Solutions/jertel/wip

refactored soc config

VERSION

@@ -1 +1 @@
-3.2.0
+3.0.0-kilo

salt/kratos/soc_kratos.yaml

@@ -103,7 +103,7 @@ kratos:
   config:
     session:
       lifespan: 
-        description: Defines the length of a login session.
+        description: Defines the length of a login session before it will timeout, and require a new login.
         global: True
         helpLink: kratos
       whoami:

salt/manager/tools/sbin/soup

@@ -370,8 +370,9 @@ preupgrade_changes() {
     # This function is to add any new pillar items if needed.
     echo "Checking to see if changes are needed."
 
-    [[ "$INSTALLEDVERSION" =~ ^2\.4\.21[0-9]+$ ]] && up_to_3.0.0   
+    [[ "$INSTALLEDVERSION" =~ ^2\.4\.21[0-9]+$ ]] && up_to_3.0.0
     [[ "$INSTALLEDVERSION" == "3.0.0" ]] && up_to_3.1.0
+    [[ "$INSTALLEDVERSION" == "3.1.0" ]] && up_to_3.2.0
     true
 }
 
@@ -381,6 +382,7 @@ postupgrade_changes() {
 
     [[ "$POSTVERSION" =~ ^2\.4\.21[0-9]+$ ]] && post_to_3.0.0
     [[ "$POSTVERSION" == "3.0.0" ]] && post_to_3.1.0
+    [[ "$POSTVERSION" == "3.1.0" ]] && post_to_3.2.0
     true
 }
 
@@ -742,6 +744,42 @@ post_to_3.1.0() {
 
 ### 3.1.0 End ###
 
+### 3.2.0 Scripts ###
+
+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."
+  # 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
+  # so psql inside init-db.sh resolves the connect user correctly.
+  local exec_cmd="docker exec -u postgres -e POSTGRES_USER=postgres so-postgres bash /docker-entrypoint-initdb.d/init-db.sh"
+  if ! /usr/sbin/so-postgres-wait; then
+    FINAL_MESSAGE_QUEUE+=("WARNING: so-postgres was not ready during the 3.2.0 upgrade; the so_soc database may not have been bootstrapped. Re-run manually: $exec_cmd")
+    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")
+    return 0
+  fi
+  echo "so_soc bootstrap complete."
+}
+
+up_to_3.2.0() {
+  INSTALLEDVERSION=3.2.0
+}
+
+post_to_3.2.0() {
+  bootstrap_so_soc_database
+
+  POSTVERSION=3.2.0
+}
+
+### 3.2.0 End ###
+
 
 repo_sync() {
   echo "Sync the local repo."

salt/postgres/files/init-db.sh

@@ -32,3 +32,8 @@ EOSQL
 if ! psql -U "$POSTGRES_USER" -tAc "SELECT 1 FROM pg_database WHERE datname='so_telegraf'" | grep -q 1; then
     psql -v ON_ERROR_STOP=1 -U "$POSTGRES_USER" -c "CREATE DATABASE so_telegraf"
 fi
+
+# Bootstrap the SOC database.
+if ! psql -U "$POSTGRES_USER" -tAc "SELECT 1 FROM pg_database WHERE datname='so_soc'" | grep -q 1; then
+    psql -v ON_ERROR_STOP=1 -U "$POSTGRES_USER" -c "CREATE DATABASE so_soc"
+fi

salt/postgres/telegraf_users.sls

@@ -18,26 +18,12 @@ include:
 {% set TG_OUT = TELEGRAFMERGED.output | upper %}
 {% if TG_OUT in ['POSTGRES', 'BOTH'] %}
 
-# docker_container.running returns as soon as the container starts, but on
-# first-init docker-entrypoint.sh starts a temporary postgres with
-# `listen_addresses=''` to run /docker-entrypoint-initdb.d scripts, then
-# shuts it down before exec'ing the real CMD. A default pg_isready check
-# (Unix socket) passes during that ephemeral phase and races the shutdown
-# with "the database system is shutting down". Checking TCP readiness on
-# 127.0.0.1 only succeeds after the final postgres binds the port.
 postgres_wait_ready:
   cmd.run:
-    - name: |
-        for i in $(seq 1 60); do
-          if docker exec so-postgres pg_isready -h 127.0.0.1 -U postgres -q 2>/dev/null; then
-            exit 0
-          fi
-          sleep 2
-        done
-        echo "so-postgres did not accept TCP connections within 120s" >&2
-        exit 1
+    - name: /usr/sbin/so-postgres-wait
     - require:
       - docker_container: so-postgres
+      - file: postgres_sbin
 
 # Ensure the shared Telegraf database exists. init-db.sh only runs on a
 # fresh data dir, so hosts upgraded onto an existing /nsm/postgres volume

salt/postgres/tools/sbin/so-postgres-wait

@@ -0,0 +1,32 @@
+#!/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.
+
+# Wait for the so-postgres container to accept TCP connections.
+#
+# docker_container.running returns as soon as the container starts, but on
+# first-init docker-entrypoint.sh starts a temporary postgres with
+# `listen_addresses=''` to run /docker-entrypoint-initdb.d scripts, then
+# shuts it down before exec'ing the real CMD. A default pg_isready check
+# (Unix socket) passes during that ephemeral phase and races the shutdown
+# with "the database system is shutting down". Checking TCP readiness on
+# 127.0.0.1 only succeeds after the final postgres binds the port.
+#
+# Usage: so-postgres-wait [iterations] [sleep_seconds]
+# Default: 60 iterations, 2s sleep (~120s total).
+
+ITERATIONS=${1:-60}
+SLEEP_SECONDS=${2:-2}
+
+for i in $(seq 1 "$ITERATIONS"); do
+  if docker exec so-postgres pg_isready -h 127.0.0.1 -U postgres -q 2>/dev/null; then
+    exit 0
+  fi
+  sleep "$SLEEP_SECONDS"
+done
+
+echo "so-postgres did not accept TCP connections within $((ITERATIONS * SLEEP_SECONDS))s" >&2
+exit 1

salt/soc/defaults.yaml

@@ -1519,6 +1519,12 @@ soc:
               serviceAccountJSON: ""
               serviceAccountLocation: ""
               healthTimeoutSeconds: 5
+        onionconfig:
+          saltstackDir: /opt/so/saltstack
+          bypassEnabled: false
+        postgres:
+          host:
+          password:
         salt:
           queueDir: /opt/sensoroni/queue
           timeoutMs: 45000

salt/soc/merged.map.jinja

@@ -16,6 +16,13 @@
 {% do SOCMERGED.config.server.update({'additionalCA': MANAGERMERGED.additionalCA}) %}
 {% do SOCMERGED.config.server.update({'insecureSkipVerify': MANAGERMERGED.insecureSkipVerify}) %}
 
+{% if not SOCMERGED.config.server.modules.postgres.host %}
+{%   do SOCMERGED.config.server.modules.postgres.update({'host': GLOBALS.manager}) %}
+{% endif %}
+{% if not SOCMERGED.config.server.modules.postgres.password %}
+{%   do SOCMERGED.config.server.modules.postgres.update({'password': salt['pillar.get']('secrets:postgres_pass', '')}) %}
+{% endif %}
+
 {# if SOCMERGED.config.server.modules.cases == httpcase details come from the soc pillar #}
 {% if SOCMERGED.config.server.modules.cases != 'soc' %}
 {%   do SOCMERGED.config.server.modules.elastic.update({'casesEnabled': false}) %}

salt/soc/soc_soc.yaml

@@ -453,6 +453,26 @@ soc:
             description: Duration (in milliseconds) that must elapse after a grid node fails to check-in before the node will be marked offline (fault).
             global: True
             advanced: True
+        onionconfig:
+          saltstackDir:
+            description: Root directory containing the SaltStack tree that SOC reads and writes configuration from. Should not be changed under normal circumstances.
+            global: True
+            advanced: True
+          bypassEnabled:
+            description: When enabled, errors encountered while reading the SaltStack pillar tree (missing files, unreadable directories, etc.) are logged but do not prevent SOC from starting or serving settings. Intended for advanced troubleshooting and recovery scenarios when the pillar tree is partially unreadable.
+            global: True
+            advanced: True
+            forcedType: bool
+        postgres:
+          host:
+            description: Hostname or IP address of the PostgreSQL server used by SOC. Defaults to the manager hostname.
+            global: True
+            advanced: True
+          password:
+            description: Password used by SOC to authenticate to the PostgreSQL server. Defaults to the postgres superuser password seeded in the secrets pillar.
+            global: True
+            sensitive: True
+            advanced: True
         salt:
           longRelayTimeoutMs:
             description: Duration (in milliseconds) to wait for a response from the Salt API when executing tasks known for being long running before giving up and showing an error on the SOC UI.
@@ -818,6 +838,7 @@ soc:
           description: List of available external tools visible in the SOC UI. Each tool is defined in JSON object notation, and must include the "name" key and "link" key, where the link is the tool's URL.
           global: True
           advanced: True
+          multiline: True
           forcedType: "[]{}"
         exportNodeId:
           description: The node ID on which export jobs will be executed.