Remove dead PCIe hardware at array start so correctable AER errors stop burying the syslog

Deployment/master.conf.template

@@ -358,6 +358,7 @@
 # Watchdogs (resource_watchdog, docker_watchdog, system_watchdog) are cronned via
 # watchdog_orchestrator.sh — NOT launched here.
     ARRAY_START_SCRIPTS=(
+        "Plugin/unraid/System_Essentials/pcie_aer_quiet.sh"         # drop AER-spamming dead hardware — runs first so later logs stay readable
         "Plugin/unraid/System_Essentials/unraid_api_key_renew.sh"   # re-register Varaverk API key — registry is ephemeral
         "System_Essentials/conf_sync.sh"              # pull partner confs + push own conf into /tmp/.vv/ RAM cache
         "System_Essentials/conf_cache_restore.sh"     # load partner confs from persistent backup if conf_sync couldn't reach partner
@@ -993,6 +994,19 @@
 # log lines per hour that have no diagnostic value. Filter removes them at source.
     FILTER_FILE="/etc/rsyslog.d/ignore-docker-veth.conf"
 
+# ━━━ PCIe AER Quiet ━━━
+# Master gate for pcie_aer_quiet.sh — removes dead PCIe hardware from the bus at
+# array start so it stops flooding syslog with correctable AER errors.
+#
+# Correctable means the link recovered, so the errors are harmless — but the kernel
+# logs every one. Removing the device ends it at the source. Unlike pci=noaer this
+# keeps uncorrectable AER reporting alive on every other device, which the AI repair
+# triage relies on to tell a real fault from this noise.
+#
+# Devices are listed per host in host*.conf as HOST*_PCIE_QUIET_DEVICES.
+# Off by default — turn on only after filling in that list for this host.
+    PCIE_QUIET_ENABLED=false
+
 # ━━━ PHP-FPM ━━━
 # Higher max_children allows more concurrent PHP requests to the unRAID WebGUI.
 # Default is very low — increasing it prevents WebGUI slowdowns under load.