merge 3/dev

salt/common/init.sls

@@ -141,6 +141,20 @@ pin_nic_names:
       - file: common_sbin
       - file: statedir
 
+# Once a node is actually running UEK8, the stock EL9 (RHCK) kernel packages are dead weight.
+# They can't be removed any earlier -- dnf protects the running kernel -- so the cleanup waits
+# for the reboot, which makes the highstate the natural place to catch it: fresh installs
+# reboot at the end of setup, and upgraded nodes reboot whenever the admin schedules it.
+# so-kernel-upgrade --cleanup checks rpm before touching dnf, so this costs an rpm query on
+# every highstate after the first pass. The package list lives in the script only, so there
+# is nothing here to drift out of sync with it.
+remove_stock_kernel:
+  cmd.run:
+    - name: /usr/sbin/so-kernel-upgrade --cleanup
+    - onlyif: 'uname -r | grep -qE "^6\.[0-9]+.*uek"'
+    - require:
+      - file: common_sbin
+
 common_sbin_jinja:
   file.recurse:
     - name: /usr/sbin

salt/common/tools/sbin/so-kernel-upgrade

@@ -5,10 +5,11 @@
 # https://securityonion.net/license; you may not use this file except in compliance with the
 # Elastic License 2.0.
 #
-# so-kernel-upgrade — install the UEK8 (6.x) kernel and make it the boot default.
+# so-kernel-upgrade — install the UEK8 (6.x) kernel, make it the boot default, and once the
+# node is running it, remove the stock EL9 kernel.
 #
 # Security Onion is moving off the EL9 stock kernel (RHCK, 5.14) and UEK7 (5.15) onto UEK8
-# (6.x). Three things have to happen, and the tool has to drive each one:
+# (6.x). Four things have to happen, and the tool has to drive each one:
 #
 #   1. Populate. The manager mirrors the UEK8 packages into /nsm/kernelrepo via so-repo-sync,
 #      and serves them to the grid over https://<manager>/kernelrepo. Until that sync runs the
@@ -26,10 +27,21 @@
 #        - From the stock EL9 kernel (RHCK, 5.14, no UEK) it is a flavor CROSS that is NOT
 #          auto-promoted, so the box keeps booting RHCK until grubby is told otherwise.
 #      This tool inspects the running kernel and only runs 'grubby --set-default' for RHCK.
+#   4. Clean up. Once the node is actually RUNNING UEK8 the stock kernel packages are dead
+#      weight -- disk in /boot and a stale GRUB entry. They cannot come off any earlier:
+#      dnf's protect_running_kernel refuses to erase the booted kernel-core, so the removal
+#      has to wait for the reboot. Waiting is also the safer sequencing on its own terms --
+#      the node has proven it comes up on UEK8 before its fallback is deleted. That is why
+#      the removal does not happen in the uek7 branch either, where dnf would allow it.
 #
 # Every one of those failure modes is silent by default. This tool handles each case and fails
 # loudly when it cannot, rather than reporting success while changing nothing.
 #
+# Invocation: with no arguments it drives the whole sequence for whatever kernel the node is
+# on. With --cleanup it does the step 4 removal ONLY, and no-ops on a node that isn't running
+# UEK8 yet -- that is the form the common highstate calls (remove_stock_kernel in
+# salt/common/init.sls) so the cleanup lands grid-wide after each node reboots.
+#
 # Manager vs minion: only the manager owns /nsm/kernelrepo, so only the manager can populate
 # it. If the repo is empty here, a manager runs so-repo-sync itself; a minion has no way to
 # fix it and exits non-zero telling the admin to sync the manager first.
@@ -49,6 +61,11 @@ KERNEL_REPO_DIR="/nsm/kernelrepo"
 REPOSYNC_CONF="/opt/so/conf/reposync/repodownload.conf"
 GLOBAL_PILLAR="/opt/so/saltstack/local/pillar/global/soc_global.sls"
 
+# Stock EL9 (RHCK) kernel packages, removed only once the node is running UEK8 (see step 4
+# in the header). Left deliberately narrow: UEK7 kernel-uek builds age out on their own via
+# installonly_limit=3, and kernel-devel/kernel-headers are not touched.
+RHCK_PKGS="kernel kernel-core kernel-modules kernel-modules-core kernel-tools kernel-tools-libs"
+
 log() { echo "[so-kernel-upgrade] $*"; }
 die() { echo "[so-kernel-upgrade] ERROR: $*" >&2; exit 1; }
 
@@ -149,8 +166,13 @@ ensure_kernel_repo() {
 }
 
 reboot_notice() {
-    [ "$(uname -r)" = "$(basename "$1" | sed 's/^vmlinuz-//')" ] \
-        || log "REBOOT REQUIRED to start using the UEK8 kernel (currently running $(uname -r))."
+    [ "$(uname -r)" = "$(basename "$1" | sed 's/^vmlinuz-//')" ] && return 0
+    log "REBOOT REQUIRED to start using the UEK8 kernel (currently running $(uname -r))."
+    # The stock kernel can't be removed until it stops being the running one, so say when
+    # that will happen rather than leaving the admin to wonder if it was missed.
+    [ -n "$(rhck_installed)" ] \
+        && log "The stock EL9 kernel is left in place until then; it is removed by the next highstate after the reboot."
+    return 0
 }
 
 # Keep future kernel updates on the UEK line rather than falling back to RHCK. Oracle ships
@@ -162,6 +184,32 @@ set_default_kernel_conf() {
     fi
 }
 
+# Which of RHCK_PKGS are actually installed, one per line. rpm -qa treats each argument as a
+# name glob and prints only what it finds, so a package that was never installed (or is
+# already gone) simply doesn't appear -- no "not installed" noise and no non-zero exit.
+rhck_installed() {
+    rpm -qa $RHCK_PKGS 2>/dev/null
+}
+
+# Remove the stock EL9 kernel. Only ever called once the running kernel is UEK8. The rpm
+# check above is the idempotency guard, so this is a cheap no-op on every highstate after
+# the first one -- it costs an rpm query, not a dnf transaction.
+remove_rhck() {
+    local installed; installed="$(rhck_installed)"
+    if [ -z "$installed" ]; then
+        log "no stock EL9 (RHCK) kernel packages installed; nothing to remove."
+        return 0
+    fi
+
+    log "running UEK8; removing the stock EL9 (RHCK) kernel packages:"
+    echo "$installed" | sed 's/^/[so-kernel-upgrade]   /'
+    dnf -y remove $RHCK_PKGS || die "failed to remove the stock EL9 kernel packages"
+
+    installed="$(rhck_installed)"
+    [ -z "$installed" ] || die "dnf reported success but these remain: $(echo $installed)"
+    log "stock EL9 kernel packages removed."
+}
+
 # Make sure a UEK8 kernel is installed, leaving its boot entry in INSTALLED_UEK8. If one is
 # already present we leave the repo alone -- it may be disabled or empty and we don't need it
 # just to flip the boot default. Otherwise install the explicit NEVRA, not the bare package
@@ -184,12 +232,38 @@ ensure_uek8_installed() {
     log "installed UEK8 kernel: $INSTALLED_UEK8"
 }
 
+# --cleanup does step 4 and nothing else. It exits 0 rather than failing on a node that
+# isn't on UEK8 yet: the highstate gates on 'uname -r' before calling this, and a state that
+# fails whenever that gate races would be worse than one that says what it's waiting for.
+case "$1" in
+"")
+    ;;
+--cleanup)
+    if [ "$(running_flavor)" != uek8 ]; then
+        log "not running a UEK8 kernel yet (currently $(uname -r)); leaving the stock EL9 kernel in place."
+        log "Run so-kernel-upgrade with no arguments to install UEK8, then reboot."
+        exit 0
+    fi
+    set_default_kernel_conf
+    remove_rhck
+    exit 0
+    ;;
+*)
+    echo "Usage: so-kernel-upgrade [--cleanup]" >&2
+    echo "  (no arguments)  install UEK8, make it the boot default, clean up once it's running" >&2
+    echo "  --cleanup       remove the stock EL9 kernel; no-op unless already running UEK8" >&2
+    exit 1
+    ;;
+esac
+
 case "$(running_flavor)" in
 uek8)
     # Already on the 6.x UEK line. A plain 'dnf update' keeps this node current within the
-    # lineage and auto-promotes newer builds, so there is nothing for this tool to do.
-    log "already running a UEK8 kernel ($(uname -r)); nothing to do."
-    exit 0
+    # lineage and auto-promotes newer builds, so there is no install or grubby work left --
+    # only the step 4 cleanup, which this is the first point in the sequence that can run it.
+    log "already running a UEK8 kernel ($(uname -r)); no kernel install needed."
+    set_default_kernel_conf
+    remove_rhck
     ;;
 
 uek7)

salt/logstash/config.sls

@@ -81,6 +81,14 @@ ls_custom_pipeline_conf_{{assigned_pipeline}}_{{pipeline}}:
 
 
 {% for assigned_pipeline in ASSIGNED_PIPELINES %}
+{# a blank per-pipeline setting falls back to the global logstash.yml value #}
+{% set PARSED_OVERRIDES = LOGSTASH_MERGED.get('pipeline_settings', {}).get(assigned_pipeline, {}) %}
+{% if PARSED_OVERRIDES is not mapping %}
+{%   do salt.log.warning('logstash: ignoring malformed pipeline_settings for pipeline ' ~ assigned_pipeline ~ '; expected a set of settings') %}
+{% endif %}
+{% set PIPELINE_OVERRIDES = PARSED_OVERRIDES if PARSED_OVERRIDES is mapping else {} %}
+{% set THREADS = PIPELINE_OVERRIDES.get('pipeline_x_workers') or LOGSTASH_MERGED.config.pipeline_x_workers %}
+{% set BATCH = PIPELINE_OVERRIDES.get('pipeline_x_batch_x_size') or LOGSTASH_MERGED.config.pipeline_x_batch_x_size %}
     {% for CONFIGFILE in LOGSTASH_MERGED.defined_pipelines[assigned_pipeline] %}
 ls_pipeline_{{assigned_pipeline}}_{{CONFIGFILE.split('.')[0] | replace("/","_") }}:
   file.managed:
@@ -92,8 +100,8 @@ ls_pipeline_{{assigned_pipeline}}_{{CONFIGFILE.split('.')[0] | replace("/","_")
         GLOBALS: {{ GLOBALS }}
         ES_USER: "{{ salt['pillar.get']('elasticsearch:auth:users:so_elastic_user:user', '') }}"
         ES_PASS: "{{ salt['pillar.get']('elasticsearch:auth:users:so_elastic_user:pass', '') }}"
-        THREADS: {{ LOGSTASH_MERGED.config.pipeline_x_workers }}
-        BATCH: {{ LOGSTASH_MERGED.config.pipeline_x_batch_x_size }}
+        THREADS: {{ THREADS }}
+        BATCH: {{ BATCH }}
       {% else %}
     - name: /opt/so/conf/logstash/pipelines/{{assigned_pipeline}}/{{CONFIGFILE.split('/')[1]}}
       {% endif %}
@@ -125,6 +133,14 @@ lspipelinesyml:
     - defaults:
         ASSIGNED_PIPELINES: {{ ASSIGNED_PIPELINES }}
 
+lslog4j2:
+  file.managed:
+    - name: /opt/so/conf/logstash/etc/log4j2.properties
+    - source: salt://logstash/etc/log4j2.properties.jinja
+    - template: jinja
+    - user: 931
+    - group: 939
+
 lsetcsync:
   file.recurse:
     - name: /opt/so/conf/logstash/etc
@@ -133,7 +149,11 @@ lsetcsync:
     - group: 939
     - template: jinja
     - clean: True
-    - exclude_pat: pipelines*
+{#- both names are matched: the .jinja source so the recurse does not copy it verbatim,
+    and the rendered file so clean: True does not delete what lslog4j2 wrote #}
+    - exclude_pat:
+      - pipelines*
+      - log4j2.properties*
     - defaults:
         LOGSTASH_MERGED: {{ LOGSTASH_MERGED }}
 

salt/logstash/defaults.yaml

@@ -42,6 +42,11 @@ logstash:
     custom2: []
     custom3: []
     custom4: []
+    custom5: []
+    custom6: []
+    custom7: []
+    custom8: []
+    custom9: []
   pipeline_config:
     custom001: |-
       filter {
@@ -60,10 +65,405 @@ logstash:
     custom008: PLACEHOLDER
     custom009: PLACEHOLDER
     custom010: PLACEHOLDER
+  pipeline_settings:
+    fleet:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    manager:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    receiver:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    search:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    custom0:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    custom1:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    custom2:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    custom3:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    custom4:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    custom5:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    custom6:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    custom7:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    custom8:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
+    custom9:
+      pipeline_x_workers: ''
+      pipeline_x_batch_x_size: ''
+      pipeline_x_batch_x_delay: ''
+      pipeline_x_batch_x_metrics_x_sampling_mode: ''
+      pipeline_x_ordered: ''
+      pipeline_x_ecs_compatibility: ''
+      pipeline_x_reloadable: ''
+      queue_x_type: ''
+      queue_x_max_bytes: ''
+      queue_x_page_capacity: ''
+      queue_x_max_events: ''
+      queue_x_checkpoint_x_acks: ''
+      queue_x_checkpoint_x_writes: ''
+      queue_x_checkpoint_x_interval: ''
+      queue_x_checkpoint_x_retry: ''
+      queue_x_compression: ''
+      queue_x_drain: ''
+      dead_letter_queue_x_enable: ''
+      dead_letter_queue_x_max_bytes: ''
+      dead_letter_queue_x_flush_interval: ''
+      dead_letter_queue_x_flush_check_interval: ''
+      dead_letter_queue_x_storage_policy: ''
+      dead_letter_queue_x_retain_x_age: ''
+      path_x_queue: ''
+      path_x_dead_letter_queue: ''
+      config_x_debug: ''
+      config_x_support_escapes: ''
   settings:
     lsheap: 500m
   config:
     api_x_http_x_host: 0.0.0.0
+    log_x_level: info
+    log_x_format: plain
     path_x_logs: /var/log/logstash
     pipeline_x_workers: 1
     pipeline_x_batch_x_size: 125

salt/logstash/enabled.sls

@@ -105,6 +105,8 @@ so-logstash:
     {% endif %}
     - watch:
       - file: lsetcsync
+      - file: lslog4j2
+      - file: lspipelinesyml
       - file: trusttheca
       {% if GLOBALS.is_manager %}
       - file: elasticsearch_cacerts

salt/logstash/etc/log4j2.properties.jinja

@@ -1,3 +1,4 @@
+{%- from 'logstash/map.jinja' import LOGSTASH_MERGED -%}
 status = error
 name = LogstashPropertiesConfig
 
@@ -16,8 +17,14 @@ name = LogstashPropertiesConfig
 appender.rolling.type = RollingFile
 appender.rolling.name = rolling
 appender.rolling.fileName = /var/log/logstash/logstash.log
+{%- if LOGSTASH_MERGED.config.get('log_x_format', 'plain') == 'json' %}
+appender.rolling.layout.type = JSONLayout
+appender.rolling.layout.compact = true
+appender.rolling.layout.eventEol = true
+{%- else %}
 appender.rolling.layout.type = PatternLayout
 appender.rolling.layout.pattern = [%d{ISO8601}][%-5p][%-25c] %.10000m%n
+{%- endif %}
 appender.rolling.filePattern = /var/log/logstash/logstash-%d{yyyy-MM-dd}.log.gz
 appender.rolling.policies.type = Policies
 appender.rolling.policies.time.type = TimeBasedTriggeringPolicy
@@ -32,7 +39,5 @@ appender.rolling.strategy.action.condition.type = IfFileName
 appender.rolling.strategy.action.condition.glob = *.gz
 appender.rolling.strategy.action.condition.nested_condition.type = IfLastModified
 appender.rolling.strategy.action.condition.nested_condition.age = 7D
-rootLogger.level = info
+rootLogger.level = ${sys:ls.log.level}
 rootLogger.appenderRef.rolling.ref = rolling
-#rootLogger.level = ${sys:ls.log.level}
-#rootLogger.appenderRef.console.ref = ${sys:ls.log.format}_console

salt/logstash/etc/pipelines.yml.jinja

@@ -1,4 +1,17 @@
+{%- from 'logstash/map.jinja' import LOGSTASH_MERGED %}
+{%- set PIPELINE_SETTINGS = LOGSTASH_MERGED.get('pipeline_settings', {}) %}
 {%- for assigned_pipeline in ASSIGNED_PIPELINES %}
 - pipeline.id: {{ assigned_pipeline }}
   path.config: "/usr/share/logstash/pipelines/{{ assigned_pipeline }}/"
+{%-   set extra = PIPELINE_SETTINGS.get(assigned_pipeline, {}) %}
+{%-   if extra is mapping %}
+{#-     values are emitted unquoted so yaml re-infers the type logstash expects:
+        4 as an integer, false as a boolean, 1024mb and auto as strings #}
+{%-     for key, value in extra | dictsort %}
+{%-       set rendered = key | replace('_x_', '.') %}
+{%-       if value not in ['', None] and rendered not in ['pipeline.id', 'path.config'] %}
+  {{ rendered }}: {{ value }}
+{%-       endif %}
+{%-     endfor %}
+{%-   endif %}
 {% endfor -%}

salt/logstash/soc_logstash.yaml

@@ -16,6 +16,7 @@ logstash:
       heavynode: *assigned_pipelines
       searchnode: *assigned_pipelines
       manager: *assigned_pipelines
+      managerhype: *assigned_pipelines
       managersearch: *assigned_pipelines
       fleet: *assigned_pipelines
   defined_pipelines:
@@ -34,6 +35,11 @@ logstash:
     custom2: *defined_pipelines
     custom3: *defined_pipelines
     custom4: *defined_pipelines
+    custom5: *defined_pipelines
+    custom6: *defined_pipelines
+    custom7: *defined_pipelines
+    custom8: *defined_pipelines
+    custom9: *defined_pipelines
   pipeline_config:
     custom001: &pipeline_config
       description: Pipeline configuration for Logstash
@@ -51,6 +57,351 @@ logstash:
     custom008: *pipeline_config
     custom009: *pipeline_config
     custom010: *pipeline_config
+  pipeline_settings:
+    manager: &pipeline_settings
+      pipeline_x_workers:
+        description: >-
+          Number of worker threads that run filters and outputs for this pipeline. May be set higher
+          than the CPU core count when outputs spend time waiting on I/O. Leave blank to use the value
+          from logstash.yml.
+        title: pipeline.workers
+        regex: '^$|^[1-9][0-9]*$'
+        regexFailureMessage: Must be blank, or a positive whole number.
+        advanced: True
+        global: False
+        helpLink: logstash
+      pipeline_x_batch_x_size:
+        description: >-
+          Maximum number of events an individual worker thread collects before running filters and
+          outputs. Larger batches are more efficient but increase heap use; total in-flight events is
+          workers multiplied by batch size. Leave blank to use the value from logstash.yml.
+        title: pipeline.batch.size
+        regex: '^$|^[1-9][0-9]*$'
+        regexFailureMessage: Must be blank, or a positive whole number.
+        advanced: True
+        global: False
+        helpLink: logstash
+      pipeline_x_batch_x_delay:
+        description: >-
+          Milliseconds a worker waits for the next event before running a batch that is not yet full.
+          Leave blank to use the value from logstash.yml.
+        title: pipeline.batch.delay
+        regex: '^$|^[0-9]+$'
+        regexFailureMessage: Must be blank, or a whole number.
+        advanced: True
+        global: False
+        helpLink: logstash
+      pipeline_x_batch_x_metrics_x_sampling_mode:
+        description: >-
+          Controls how often batch size metrics are collected for this pipeline, which helps tune
+          pipeline.batch.size to the batch sizes actually being processed. Fuller sampling consumes
+          additional heap. Elastic marks this setting as a technical preview that may change in a
+          future release. Leave blank to use the value from logstash.yml.
+        title: pipeline.batch.metrics.sampling_mode
+        options:
+        - ''
+        - 'disabled'
+        - 'minimal'
+        - 'full'
+        advanced: True
+        global: False
+        helpLink: logstash
+      pipeline_x_ordered:
+        description: >-
+          Whether event order is preserved through this pipeline. auto enables ordering only when
+          pipeline.workers is explicitly set to 1, and does nothing otherwise. Setting this to true
+          requires pipeline.workers to be 1 as well; with more workers this pipeline fails to start.
+          Leave blank to use the value from logstash.yml.
+        title: pipeline.ordered
+        options:
+        - ''
+        - 'auto'
+        - 'true'
+        - 'false'
+        advanced: True
+        global: False
+        helpLink: logstash
+      pipeline_x_ecs_compatibility:
+        description: >-
+          Elastic Common Schema compatibility mode for plugins in this pipeline. Security Onion sets
+          this globally and it should rarely be changed per pipeline. Elastic considers values other
+          than disabled to be BETA, and they may produce unintended consequences when upgrading
+          Logstash. Leave blank to use the value from logstash.yml.
+        title: pipeline.ecs_compatibility
+        options:
+        - ''
+        - 'disabled'
+        - 'v1'
+        - 'v8'
+        advanced: True
+        global: False
+        helpLink: logstash
+      pipeline_x_reloadable:
+        description: >-
+          Whether this pipeline may be reloaded when its configuration changes. Leave blank to use the
+          value from logstash.yml.
+        title: pipeline.reloadable
+        options:
+        - ''
+        - 'true'
+        - 'false'
+        advanced: True
+        global: False
+        helpLink: logstash
+      queue_x_type:
+        description: >-
+          Queue backing this pipeline. persisted buffers events to disk under /nsm/logstash so they
+          survive a restart, at some throughput cost; memory does not. Leave blank to use the value
+          from logstash.yml.
+        title: queue.type
+        options:
+        - ''
+        - 'memory'
+        - 'persisted'
+        advanced: True
+        global: False
+        helpLink: logstash
+      queue_x_max_bytes:
+        description: >-
+          Total capacity of the persistent queue for this pipeline, in bytes. Only applies when
+          queue.type is persisted. The disk backing /nsm/logstash must have room for this much data or
+          the pipeline fails to start, reporting that it was unable to allocate the space. If both
+          queue.max_events and queue.max_bytes are set, whichever is reached first applies. Leave
+          blank to use the value from logstash.yml.
+        title: queue.max_bytes
+        regex: '^$|^[0-9]+$|^[0-9]+(\.[0-9]+)?\s*(b|kb?|mb?|gb?|tb?|pb?)$'
+        regexFailureMessage: Must be blank, or a size such as 512mb, 1gb, or 64k. Units are lowercase.
+        advanced: True
+        global: False
+        helpLink: logstash
+      queue_x_page_capacity:
+        description: >-
+          Size of the individual append-only page data files that make up the persistent queue for
+          this pipeline. Only applies when queue.type is persisted. Leave blank to use the value from
+          logstash.yml.
+        title: queue.page_capacity
+        regex: '^$|^[0-9]+$|^[0-9]+(\.[0-9]+)?\s*(b|kb?|mb?|gb?|tb?|pb?)$'
+        regexFailureMessage: Must be blank, or a size such as 512mb, 1gb, or 64k. Units are lowercase.
+        advanced: True
+        global: False
+        helpLink: logstash
+      queue_x_max_events:
+        description: >-
+          Maximum number of unread events in the persistent queue for this pipeline. 0 means
+          unlimited. Only applies when queue.type is persisted. Leave blank to use the value from
+          logstash.yml.
+        title: queue.max_events
+        regex: '^$|^[0-9]+$'
+        regexFailureMessage: Must be blank, or a whole number.
+        advanced: True
+        global: False
+        helpLink: logstash
+      queue_x_checkpoint_x_acks:
+        description: >-
+          Maximum number of acknowledged events before a checkpoint is forced. 0 means unlimited. Only
+          applies when queue.type is persisted. Leave blank to use the value from logstash.yml.
+        title: queue.checkpoint.acks
+        regex: '^$|^[0-9]+$'
+        regexFailureMessage: Must be blank, or a whole number.
+        advanced: True
+        global: False
+        helpLink: logstash
+      queue_x_checkpoint_x_writes:
+        description: >-
+          Maximum number of written events before a checkpoint is forced. Setting this to 1 gives
+          maximum durability at a severe performance cost. 0 means unlimited. Only applies when
+          queue.type is persisted. Leave blank to use the value from logstash.yml.
+        title: queue.checkpoint.writes
+        regex: '^$|^[0-9]+$'
+        regexFailureMessage: Must be blank, or a whole number.
+        advanced: True
+        global: False
+        helpLink: logstash
+      queue_x_checkpoint_x_interval:
+        description: >-
+          Milliseconds between forced checkpoints on the persistent queue head page. 0 eliminates
+          periodic checkpoints. Deprecated by Elastic as of Logstash 9.1. Only applies when queue.type
+          is persisted. Leave blank to use the value from logstash.yml.
+        title: queue.checkpoint.interval
+        regex: '^$|^[0-9]+$'
+        regexFailureMessage: Must be blank, or a whole number.
+        advanced: True
+        global: False
+        helpLink: logstash
+      queue_x_checkpoint_x_retry:
+        description: >-
+          When enabled, Logstash retries four times per attempted checkpoint write that fails; later
+          errors are not retried. Elastic describes this as a workaround for failed checkpoint writes
+          seen only on Windows and on filesystems with non-standard behaviour such as SANs, and does
+          not recommend enabling it otherwise. Only applies when queue.type is persisted. Leave blank
+          to use the value from logstash.yml.
+        title: queue.checkpoint.retry
+        options:
+        - ''
+        - 'true'
+        - 'false'
+        advanced: True
+        global: False
+        helpLink: logstash
+      queue_x_compression:
+        description: >-
+          Compression applied to persistent queue pages for this pipeline, trading CPU for disk: speed
+          favours the fastest operation, size the smallest files, and balanced sits between them. Once
+          compressed events have been written, that queue cannot be read by Logstash releases earlier
+          than 9.2. Only applies when queue.type is persisted. Leave blank to use the value from
+          logstash.yml.
+        title: queue.compression
+        options:
+        - ''
+        - 'none'
+        - 'speed'
+        - 'balanced'
+        - 'size'
+        advanced: True
+        global: False
+        helpLink: logstash
+      queue_x_drain:
+        description: >-
+          When enabled, Logstash waits for the persistent queue to drain before shutting down this
+          pipeline. Draining a large queue makes shutdown take considerably longer. Only applies when
+          queue.type is persisted. Leave blank to use the value from logstash.yml.
+        title: queue.drain
+        options:
+        - ''
+        - 'true'
+        - 'false'
+        advanced: True
+        global: False
+        helpLink: logstash
+      dead_letter_queue_x_enable:
+        description: >-
+          Whether events this pipeline cannot process are written to a dead letter queue instead of
+          being dropped. Leave blank to use the value from logstash.yml.
+        title: dead_letter_queue.enable
+        options:
+        - ''
+        - 'true'
+        - 'false'
+        advanced: True
+        global: False
+        helpLink: logstash
+      dead_letter_queue_x_max_bytes:
+        description: >-
+          Total capacity of the dead letter queue for this pipeline, in bytes. Only applies when
+          dead_letter_queue.enable is true. Leave blank to use the value from logstash.yml.
+        title: dead_letter_queue.max_bytes
+        regex: '^$|^[0-9]+$|^[0-9]+(\.[0-9]+)?\s*(b|kb?|mb?|gb?|tb?|pb?)$'
+        regexFailureMessage: Must be blank, or a size such as 512mb, 1gb, or 64k. Units are lowercase.
+        advanced: True
+        global: False
+        helpLink: logstash
+      dead_letter_queue_x_flush_interval:
+        description: >-
+          Milliseconds before an incomplete dead letter queue segment is flushed and made available to
+          the dead_letter_queue input. Lower values write more, smaller segment files; higher values
+          add latency before events can be read. Only applies when dead_letter_queue.enable is true.
+          Leave blank to use the value from logstash.yml.
+        title: dead_letter_queue.flush_interval
+        regex: '^$|^[0-9]+$'
+        regexFailureMessage: Must be blank, or a whole number.
+        advanced: True
+        global: False
+        helpLink: logstash
+      dead_letter_queue_x_flush_check_interval:
+        description: >-
+          Milliseconds between checks for a stale dead letter queue segment needing a flush. Cannot be
+          set lower than 1000. Smaller values rotate segments sooner at the cost of CPU. Only applies
+          when dead_letter_queue.enable is true. Leave blank to use the value from logstash.yml.
+        title: dead_letter_queue.flush_check_interval
+        regex: '^$|^[0-9]+$'
+        regexFailureMessage: Must be blank, or a whole number.
+        advanced: True
+        global: False
+        helpLink: logstash
+      dead_letter_queue_x_storage_policy:
+        description: >-
+          Action taken when dead_letter_queue.max_bytes is reached: drop_newer stops accepting new
+          events, drop_older removes the oldest events to make room. Only applies when
+          dead_letter_queue.enable is true. Leave blank to use the value from logstash.yml.
+        title: dead_letter_queue.storage_policy
+        options:
+        - ''
+        - 'drop_newer'
+        - 'drop_older'
+        advanced: True
+        global: False
+        helpLink: logstash
+      dead_letter_queue_x_retain_x_age:
+        description: >-
+          How long an event is kept in the dead letter queue before Logstash removes it, such as 5d.
+          Units are d, h, m and s; there is no default unit, so one must be given. Only applies when
+          dead_letter_queue.enable is true. Leave blank to use the value from logstash.yml.
+        title: dead_letter_queue.retain.age
+        regex: '^$|^[0-9]+\s*[dhms]$'
+        regexFailureMessage: Must be blank, or a number followed by d, h, m, or s, such as 5d.
+        advanced: True
+        global: False
+        helpLink: logstash
+      path_x_queue:
+        description: >-
+          Directory inside the Logstash container holding the persistent queue for this pipeline. The
+          default lives under the /nsm/logstash bind mount; a path outside it will not survive a
+          container restart. Logstash creates the directory if it is missing, requires it to be
+          writable, and refuses to start if the path is a symlink. Only applies when queue.type is
+          persisted. Leave blank to use the value from logstash.yml.
+        title: path.queue
+        advanced: True
+        global: False
+        helpLink: logstash
+      path_x_dead_letter_queue:
+        description: >-
+          Directory inside the Logstash container holding the dead letter queue for this pipeline. The
+          default lives under the /nsm/logstash bind mount; a path outside it will not survive a
+          container restart. Logstash creates the directory if it is missing, requires it to be
+          writable, and refuses to start if the path is a symlink. Only applies when
+          dead_letter_queue.enable is true. Leave blank to use the value from logstash.yml.
+        title: path.dead_letter_queue
+        advanced: True
+        global: False
+        helpLink: logstash
+      config_x_debug:
+        description: >-
+          Whether the fully compiled configuration for this pipeline is written to the log. The output
+          may contain sensitive values from the pipeline configuration. Leave blank to use the value
+          from logstash.yml.
+        title: config.debug
+        options:
+        - ''
+        - 'true'
+        - 'false'
+        advanced: True
+        global: False
+        helpLink: logstash
+      config_x_support_escapes:
+        description: >-
+          Whether escape sequences such as \n and \t in this pipeline's quoted strings are
+          interpreted. Leave blank to use the value from logstash.yml.
+        title: config.support_escapes
+        options:
+        - ''
+        - 'true'
+        - 'false'
+        advanced: True
+        global: False
+        helpLink: logstash
+    fleet: *pipeline_settings
+    receiver: *pipeline_settings
+    search: *pipeline_settings
+    custom0: *pipeline_settings
+    custom1: *pipeline_settings
+    custom2: *pipeline_settings
+    custom3: *pipeline_settings
+    custom4: *pipeline_settings
+    custom5: *pipeline_settings
+    custom6: *pipeline_settings
+    custom7: *pipeline_settings
+    custom8: *pipeline_settings
+    custom9: *pipeline_settings
   settings:
     lsheap: 
       description: Heap size to use for logstash
@@ -62,6 +413,35 @@ logstash:
       helpLink: logstash
       readonly: True
       advanced: True
+    log_x_level:
+      description: >-
+        Verbosity of the Logstash log at /opt/so/log/logstash/logstash.log. debug and trace produce
+        a very large volume of log data on a busy node and should be used only while troubleshooting;
+        the log rotates at 1GB and rotated files are deleted after 7 days. Setting this to debug is
+        also what makes the per-pipeline config.debug setting emit anything.
+      title: log.level
+      options:
+      - 'fatal'
+      - 'error'
+      - 'warn'
+      - 'info'
+      - 'debug'
+      - 'trace'
+      advanced: True
+      global: False
+      helpLink: logstash
+    log_x_format:
+      description: >-
+        Layout of the Logstash log. plain writes human readable lines; json writes one JSON object
+        per line, which is easier to parse but harder to read directly. The file name and location
+        do not change.
+      title: log.format
+      options:
+      - 'plain'
+      - 'json'
+      advanced: True
+      global: False
+      helpLink: logstash
     path_x_logs: 
       description: Path inside the container to wrote logs.
       helpLink: logstash

salt/manager/tools/sbin/soup

@@ -24,10 +24,7 @@ else
   POSTVERSION=$INSTALLEDVERSION
 fi
 INSTALLEDSALTVERSION=$(salt --versions-report | grep Salt: | awk '{print $2}')
-# Optional -b override for the grid highstate batch size (a count like "5" or a
-# percentage like "25%"). Empty means so-soup-grid-highstate uses the salt:auto_apply:batch
-# pillar default.
-BATCHSIZE=
+BATCHSIZE=5
 SOUP_LOG=/root/soup.log
 SOUP_DEBUG_LOG=/root/soup-debug.log
 WHATWOULDYOUSAYYAHDOHERE=soup
@@ -455,31 +452,6 @@ highstate() {
   salt-call state.highstate -l info queue=True
 }
 
-push_grid_highstate() {
-  # Drive a batched, role-tiered highstate across the rest of the grid so remote minions
-  # pick up this upgrade now instead of waiting up to ~2.5 hours for their own scheduled
-  # highstate (the schedule moved from 15 to 120 minutes). so-soup-grid-highstate does the work
-  # via orch.push_batch; it only exists once the manager highstate has deployed this
-  # version's sbin files, so guard on it. Launch fully detached (setsid) so it survives an
-  # SSH drop, and never let it affect soup's exit status -- it is best-effort with the
-  # scheduled highstate as backstop.
-  if [[ ! -x /usr/sbin/so-soup-grid-highstate ]]; then
-    echo "so-soup-grid-highstate not present; remote nodes will converge on their scheduled highstate."
-    return 0
-  fi
-
-  local extra_args=()
-  if [[ $SALTUPGRADED == true || $UPGRADESALT -eq 1 ]]; then
-    extra_args+=(--salt-upgraded)
-  fi
-  if [[ -n "$BATCHSIZE" ]]; then
-    extra_args+=(--batch "$BATCHSIZE")
-  fi
-
-  echo "Dispatching a grid-wide highstate to remote nodes. Progress: /opt/so/log/salt/so-soup-grid-highstate"
-  setsid nohup /usr/sbin/so-soup-grid-highstate --reason soup "${extra_args[@]}" >/dev/null 2>&1 &
-}
-
 masterlock() {
   echo "Locking Salt Master"
   mv -v $TOPFILE $BACKUPTOPFILE
@@ -506,7 +478,6 @@ preupgrade_changes() {
     [[ "$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
-    [[ "$INSTALLEDVERSION" == "3.2.0" ]] && up_to_3.3.0
     true
 }
 
@@ -524,7 +495,6 @@ 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
-    [[ "$POSTVERSION" == "3.2.0" ]] && post_to_3.3.0
     # All applicable post-upgrade steps completed; clear the resume marker.
     rm -f "$POSTVERSION_FILE"
     true
@@ -907,7 +877,7 @@ post_to_3.1.0() {
 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 template1 postgres securityonion so_telegraf; do
+  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
@@ -1033,6 +1003,7 @@ post_to_3.2.0() {
 
   set_postversion 3.2.0
 }
+
 ### 3.2.0 End ###
 
 ### 3.3.0 Scripts ###
@@ -2017,9 +1988,6 @@ main() {
     # rather than reporting "already latest". The soversion/pillar writes in
     # update_version are no-ops here since the version is unchanged for a hotfix.
     update_version
-    # Push the hotfix out to the rest of the grid rather than waiting for the scheduled
-    # highstate. Hotfixes never upgrade Salt, so no --salt-upgraded pass is needed.
-    push_grid_highstate
   else
     SOUP_UPGRADE_STARTED=true
     echo ""
@@ -2186,18 +2154,13 @@ main() {
 
     if [[ $NUM_MINIONS -gt 1 ]]; then
 
-      # Actively drive the rest of the grid to this version now. The scheduled highstate
-      # runs only every 120 minutes (salt:schedule:highstate_interval_minutes), so without
-      # this remote nodes could sit on the old version for a couple of hours after soup finishes.
-      push_grid_highstate
-
       cat << EOF
 
 
 
-This appears to be a distributed deployment. soup has dispatched a batched, grid-wide highstate to update the other nodes now: Elasticsearch data nodes (search/heavynodes) first, then receivers, then sensors and the remaining nodes. Progress is logged to /opt/so/log/salt/so-soup-grid-highstate, and you can watch nodes update from the Grid section of SOC. Do not manually restart anything until you know that all the search/heavynodes in your deployment are updated. This is especially important if you are using true clustering for Elasticsearch.
+This appears to be a distributed deployment. Other nodes should update themselves at the next Salt highstate (typically within 15 minutes). Do not manually restart anything until you know that all the search/heavy nodes in your deployment are updated. This is especially important if you are using true clustering for Elasticsearch.
 
-Nodes are updated in batches, and things like network bandwidth can be a factor in how long the actual upgrade takes. If you have a heavy node on a slow link, it is going to take a while to get the containers to it. Depending on what changes happened between the versions, Elasticsearch might not be able to talk to said heavy node until the update is complete. Any node that is offline or missed a batch will converge on its own scheduled highstate (every 120 minutes by default).
+Each minion is on a random 15 minute check-in period and things like network bandwidth can be a factor in how long the actual upgrade takes. If you have a heavy node on a slow link, it is going to take a while to get the containers to it. Depending on what changes happened between the versions, Elasticsearch might not be able to talk to said heavy node until the update is complete.
 
 If it looks like you’re missing data after the upgrade, please avoid restarting services and instead make sure at least one search node has completed its upgrade. The best way to do this is to run 'sudo salt-call state.highstate' from a search node and make sure there are no errors. Typically if it works on one node it will work on the rest. Sensor nodes are less complex and will update as they check in so you can monitor those from the Grid section of SOC.
 
@@ -2237,10 +2200,8 @@ while getopts ":b:f:y" opt; do
   case ${opt} in
     b )
       BATCHSIZE="$OPTARG"
-      # Accept either a plain count (e.g. 5) or a percentage (e.g. 25%); passed through
-      # to so-soup-grid-highstate --batch, which salt's batch/batch_wait accepts in both forms.
-      if ! [[ "$BATCHSIZE" =~ ^[1-9][0-9]*%?$ ]]; then
-        echo "Batch size must be a number greater than 0, optionally with a trailing % (e.g. 5 or 25%)."
+      if ! [[ "$BATCHSIZE" =~ ^[1-9][0-9]*$ ]]; then
+        echo "Batch size must be a number greater than 0."
         exit 1
       fi
     ;;

salt/postgres/files/init-db.sh

@@ -29,6 +29,8 @@ psql -v ON_ERROR_STOP=1 --username "$POSTGRES_USER" --dbname "$POSTGRES_DB" <<-E
     -- revoking CONNECT closes the soft edge entirely.
     REVOKE CONNECT ON DATABASE "$POSTGRES_DB" FROM PUBLIC;
     GRANT CONNECT ON DATABASE "$POSTGRES_DB" TO "$SO_POSTGRES_USER";
+
+    CREATE EXTENSION IF NOT EXISTS vector;
 EOSQL
 
 # Bootstrap the Telegraf metrics database. Per-minion roles + schemas are

salt/salt/minion/boot_highstate.sls

@@ -3,6 +3,8 @@
 # https://securityonion.net/license; you may not use this file except in compliance with the
 # Elastic License 2.0.
 
+{% from 'vars/globals.map.jinja' import GLOBALS %}
+
 # Manages /etc/systemd/system/so-boot-highstate.service, a Type=oneshot
 # RemainAfterExit=yes unit that runs `salt-call state.highstate` exactly once
 # per system boot. Replaces the legacy `startup_states: highstate` minion
@@ -19,9 +21,25 @@ so_boot_highstate_unit_file:
     - onchanges_in:
       - module: systemd_reload
 
+# Non-managers never apply salt.minion during setup, so reaching this state means
+# setup is finished and the marker is safe to write unconditionally. This also
+# heals nodes installed before this fix, which have no marker and no legacy
+# startup_states line to grep for. Managers do highstate mid-setup, so they only
+# get the marker from the legacy upgrade signal; fresh installs get it from
+# mark_setup_complete in setup/so-functions.
+mark_setup_complete:
+  file.managed:
+    - name: /opt/so/state/setup-complete
+    - replace: false
+    - makedirs: True
+{% if GLOBALS.is_manager %}
+    - onlyif: "grep -qx 'startup_states: highstate' /etc/salt/minion"
+{% endif %}
+    - require_in:
+      - service: so_boot_highstate_service
+
 # Only enable once setup is complete. Until then the gate file is missing and
-# the unit's own ConditionPathExists would no-op it anyway -- this just keeps
-# `systemctl is-enabled` honest for the sync_es_users gate.
+# the unit's own ConditionPathExists would no-op it anyway.
 so_boot_highstate_service:
   service.enabled:
     - name: so-boot-highstate.service

salt/salt/minion/init.sls

@@ -87,27 +87,15 @@ set_log_levels:
 # so-boot-highstate.service (managed in salt.minion.boot_highstate), which
 # runs once per system boot only. Strip the line from /etc/salt/minion on
 # upgrade; both the commented and uncommented forms historically existed.
+# Ordered after mark_setup_complete (salt.minion.boot_highstate); the manager
+# gate there greps for this line, so it must run before we delete it.
 remove_startup_states:
   file.line:
     - name: /etc/salt/minion
     - match: 'startup_states: highstate'
     - mode: delete
-
-# Upgrade-path bridge: systems that already passed setup under the old gate
-# (`grep -x 'startup_states: highstate' /etc/salt/minion`) get a /opt/so/state/setup-complete
-# marker so so-boot-highstate.service can be enabled and the so-user_sync cron
-# in sync_es_users.sls keeps installing. Setup-in-progress systems instead get
-# the marker from `mark_setup_complete` in setup/so-functions at the right
-# moment. `replace: false` means we never overwrite a marker once written.
-mark_setup_complete_for_upgrades:
-  file.managed:
-    - name: /opt/so/state/setup-complete
-    - replace: false
-    - makedirs: True
-    - onlyif: "grep -qx 'startup_states: highstate' /etc/salt/minion"
-    - require_in:
-      - file: remove_startup_states
-      - service: so_boot_highstate_service
+    - require:
+      - file: mark_setup_complete
 
 {% endif %}
 

salt/setup/virt/setSalt.sls

@@ -8,6 +8,15 @@ set_role_grain:
     - name: role
     - value: so-{{ grains.id.split("_") | last }}
 
+# salt-cloud guests never run so-setup, so nothing else marks them setup-complete.
+# Replaces the 'startup_states: highstate' line this state used to append. No
+# GLOBALS import -- this runs before the guest's pillars exist.
+mark_setup_complete_vm_guest:
+  file.managed:
+    - name: /opt/so/state/setup-complete
+    - replace: false
+    - makedirs: True
+
 enable_salt_minion:
   service.enabled:
     - name: salt-minion

salt/soc/defaults.yaml

@@ -1537,6 +1537,18 @@ soc:
             Orchestrator: sonnet@SOAI
             Investigator: gemma@SOAI
             DetectionEngineer: gemma@SOAI
+          useMemory: true
+          useMemoryScanner: false
+          memoryScanIntervalSeconds: 300
+          memoryProximityThreshold: 0.8
+          messageProximityThreshold: 0.5
+          maxUserMemoriesToInclude: 5
+          maxGlobalMemoriesToInclude: 5
+          maxUserMemoriesToReconcile: 20
+          maxGlobalMemoriesToReconcile: 20
+          memoryModel: gemma@SOAI
+          embedModel: amazon.titan-embed-text-v2@SOAI
+          reconcileModel: gemma@SOAI
         onionconfig:
           saltstackDir: /opt/so/saltstack
           bypassEnabled: false
@@ -2671,7 +2683,7 @@ soc:
               # The id (UUIDv4) is pregenerated and can safely be used.
               # Click "Convert" to convert the Sigma rule to use Security Onion field mappings within an EQL query
               #
-              # Rule Creation Guide: https://github.com/SigmaHQ/sigma/wiki/Rule-Creation-Guide
+              # Rule Creation Guide: https://github.com/SigmaHQ/sigma/wiki/Rule-Creation-High%E2%80%90Level-Guide
               # Logsources: https://sigmahq.io/docs/basics/log-sources.html
 
               title: 'A Short Capitalized Title With Less Than 50 Characters'
@@ -2683,7 +2695,7 @@ soc:
               references:
                 - 'https://local.invalid'
               author: '@SecurityOnion'
-              date: 'YYYY/MM/DD'
+              date: '[today]'
               tags:
                 - detection.threat_hunting
                 - attack.technique_id
@@ -2727,5 +2739,14 @@ soc:
               enabled: true
               adapter: SOAI
               charsPerTokenEstimate: 4
+            - id: amazon.titan-embed-text-v2
+              displayName: amazon.titan-embed-text-v2
+              origin: USA
+              contextLimitSmall: 8192
+              contextLimitLarge: 8192
+              lowBalanceColorAlert: 500000
+              enabled: true
+              adapter: SOAI
+              charsPerTokenEstimate: 4
               
             

salt/soc/soc_soc.yaml

@@ -845,6 +845,45 @@ soc:
             DetectionEngineer:
               description: This agent manages detections and their overrides, including tuning noisy rules and authoring rule content.
               global: True
+          useMemory:
+            description: Enables the Memory system for OnionAI
+            global: True
+            forcedType: bool
+          useMemoryScanner:
+            description: Enables the memory scanner for automatic memory extraction from historical sessions.
+            global: True
+            forcedType: bool
+          memoryScanIntervalSeconds:
+            description: How long to wait in seconds between attempts to scan sessions for new memories.
+            global: True
+          memoryProximityThreshold:
+            description: Describes how close memories need to be on a floating point scale from 0.0 to 1.0 to be considered when reconciling new memories with old ones. This value is usually higher than messageProximityThreshold.
+            global: True
+          messageProximityThreshold:
+            description: Describes how close a memory needs to be to a user's message on a floating point scale from 0.0 to 1.0 to be included in the context. This value is usually lower than memoryProximityThreshold.
+            global: True
+          maxUserMemoriesToInclude:
+            description: Specify the max number of user-specific memories to include in the prompt when a user sends a message.
+            global: True
+          maxGlobalMemoriesToInclude:
+            description: Specify the max number of global memories to include in the prompt when a user sends a message.
+            global: True
+          maxUserMemoriesToReconcile:
+            description: When reconciling new user-specific memories with existing user-specific memories, this determines how many old memories may be considered.
+            global: True
+          maxGlobalMemoriesToReconcile:
+            description: When reconciling new global memories with existing global memories, this determines how many old memories may be considered.
+            global: True
+          memoryModel:
+            description: The model to use when extracting memories from sessions.
+            global: True
+          embedModel: 
+            description: The model to use when embedding a memory as a vector. Note that only memories embedded using the same model may be compared and only memories created with the model specified here will be considered when informing an agent of existing memories.
+            global: True
+            advanced: True
+          reconcileModel:
+            description: The model to use when reconciling memories that contain nearly the same content.
+            global: True
       client:
         assistant:
           enabled:

salt/stig/enabled.sls

@@ -65,6 +65,15 @@ run_remediate:
     - success_retcodes:
       - 2
 
+# Elastic Agent docker image chowns its directory to the running UID but does not
+#  chown the elastic-agent launcher symlink. fs.protected_symlinks=1 then prevents
+#  non-root users from following that launcher symlink.
+{# OSCAP rule id: xccdf_org.ssgproject.content_rule_sysctl_fs_protected_symlinks #}
+fs.protected_symlinks:
+  sysctl.present:
+    - value: 0
+    - config: /etc/sysctl.conf
+
 {# OSCAP rule id: xccdf_org.ssgproject.content_rule_disable_ctrlaltdel_burstaction #}
 disable_ctrl_alt_del_action:
   file.replace:

salt/stig/files/sos-oscap.xml

@@ -1601,7 +1601,7 @@ DISA STIG for Oracle Linux 9 V1R3.</xccdf-1.2:description>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sudoers_validate_passwd" selected="true"/>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_crypto_fips_enabled" selected="true"/>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_fs_protected_hardlinks" selected="true"/>
-        <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_fs_protected_symlinks" selected="true"/>
+        <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_fs_protected_symlinks" selected="false"/>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_kernel_core_pattern" selected="true"/>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_kernel_dmesg_restrict" selected="true"/>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_kernel_exec_shield" selected="true"/>
@@ -2202,7 +2202,7 @@ standard DISA STIG for Oracle Linux 9 profile.</xccdf-1.2:description>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sudoers_validate_passwd" selected="true"/>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_crypto_fips_enabled" selected="true"/>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_fs_protected_hardlinks" selected="true"/>
-        <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_fs_protected_symlinks" selected="true"/>
+        <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_fs_protected_symlinks" selected="false"/>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_kernel_core_pattern" selected="true"/>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_kernel_dmesg_restrict" selected="true"/>
         <xccdf-1.2:select idref="xccdf_org.ssgproject.content_rule_sysctl_kernel_exec_shield" selected="true"/>

salt/telegraf/etc/telegraf.conf

@@ -335,7 +335,7 @@
 {%-   do TELEGRAFMERGED.scripts[GLOBALS.role.split('-')[1]].remove('sostatus.sh') %}
 [[inputs.exec]]
    commands = [
-      "/scripts/sostatus.sh"
+      ["/scripts/sostatus.sh"]
    ]
    data_format = "influx"
    timeout = "15s"
@@ -346,7 +346,7 @@
 [[inputs.exec]]
    commands = [
 {%-   for script in TELEGRAFMERGED.scripts[GLOBALS.role.split('-')[1]] %}
-      "/scripts/{{script}}"{% if not loop.last %},{% endif %}
+      ["/scripts/{{script}}"]{% if not loop.last %},{% endif %}
 {%-   endfor %}
    ]
   data_format = "influx"
@@ -375,7 +375,7 @@
 {%- if GLOBALS.is_manager or GLOBALS.role == 'so-heavynode' %}
 [[ inputs.exec ]]
   commands = [
-    "/scripts/esindexsize.sh"
+    ["/scripts/esindexsize.sh"]
   ]
   data_format = "influx"
   interval = "1h"

setup/so-setup

@@ -833,6 +833,7 @@ if ! [[ -f $install_opt_file ]]; then
 		check_sos_appliance
 		drop_install_options
 		hypervisor_local_states
+		mark_setup_complete
 		verify_setup
 	fi