Fix: Scan logs in subdirectories to catch all domain errors

Issue: Was missing 500 errors from logs stored in subdirectories like
/var/log/apache2/domlogs/username/domain.com

Changed from simple glob (domlogs/*) to recursive find command that:
- Scans all files in domlogs directory AND subdirectories
- Excludes system files (bytes_log, offset, error_log, ftpxferlog, ssl_log)
- Finds ALL domain access logs regardless of location

This ensures we catch errors like "GET /ay.php HTTP/1.1" 500 that were
previously missed in subdirectory logs.

modules/website/500-error-tracker.sh

@@ -59,11 +59,12 @@ declare -A domain_user
 total_500s=0
 filtered_bots=0
 
-# Scan all domain access logs for 500 errors
+# Scan all domain access logs for 500 errors (including subdirectories)
-for log in "$DOMLOGS_DIR"/*; do
+while IFS= read -r log; do
     [ -f "$log" ] || continue
     [[ "$log" =~ (bytes_log|offset|error_log|ftpxferlog|-ssl_log)$ ]] && continue
 
+    # Extract domain from log filename
     domain="${log##*/}"
     domain="${domain%%-*}"
 
@@ -130,7 +131,7 @@ for log in "$DOMLOGS_DIR"/*; do
             echo "$domain|$user|$status|$url|$timestamp|$ip" >> "$ERRORS_500"
         fi
     done < <(tail -n 100000 "$log" 2>/dev/null)
-done
+done < <(find "$DOMLOGS_DIR" -type f ! -name "*bytes_log" ! -name "*offset*" ! -name "*error_log" ! -name "*ftpxferlog*" ! -name "*-ssl_log" 2>/dev/null)
 
 if [ "$total_500s" -eq 0 ]; then
     echo ""