Smarter backfil

This commit is contained in:
Dan Milne
2025-12-02 14:01:45 +11:00
parent 1aa77066a1
commit 032243ba6a
5 changed files with 408 additions and 34 deletions

View File

@@ -20,37 +20,71 @@ class EventsController < ApplicationController
end
def index
@events = Event.includes(:network_range, :rule).order(timestamp: :desc)
Rails.logger.debug "Found #{@events.count} total events"
Rails.logger.debug "Action: #{params[:waf_action]}"
# Build filters hash from params
filters = {}
filters[:ip] = params[:ip] if params[:ip].present?
filters[:waf_action] = params[:waf_action] if params[:waf_action].present?
filters[:country] = params[:country] if params[:country].present?
filters[:rule_id] = params[:rule_id] if params[:rule_id].present?
filters[:company] = params[:company] if params[:company].present?
filters[:network_type] = params[:network_type] if params[:network_type].present?
filters[:asn] = params[:asn] if params[:asn].present?
filters[:exclude_bots] = params[:exclude_bots] if params[:exclude_bots] == "true"
# Apply filters
@events = @events.by_ip(params[:ip]) if params[:ip].present?
@events = @events.by_waf_action(params[:waf_action]) if params[:waf_action].present?
@events = @events.by_country(params[:country]) if params[:country].present?
@events = @events.where(rule_id: params[:rule_id]) if params[:rule_id].present?
# Handle network_cidr filter (requires NetworkRange lookup)
if params[:network_cidr].present?
range = NetworkRange.find_by(network: params[:network_cidr])
filters[:network_range_id] = range.id if range
end
# Network-based filters (now using denormalized columns)
@events = @events.by_company(params[:company]) if params[:company].present?
@events = @events.by_network_type(params[:network_type]) if params[:network_type].present?
@events = @events.by_asn(params[:asn]) if params[:asn].present?
@events = @events.by_network_cidr(params[:network_cidr]) if params[:network_cidr].present?
# Try DuckDB first, fallback to PostgreSQL if unavailable
result = EventDdb.search(filters, page: params[:page]&.to_i || 1, per_page: 50)
# Bot filtering
@events = @events.exclude_bots if params[:exclude_bots] == "true"
if result
# DuckDB query succeeded
@pagy = Pagy.new(count: result[:total_count], page: result[:page], items: result[:per_page])
@events = result[:events]
Rails.logger.debug "Events count after filtering: #{@events.count}"
# Load network_range associations for events that have network_range_id
network_range_ids = @events.map(&:network_range_id).compact.uniq
if network_range_ids.any?
network_ranges = NetworkRange.where(id: network_range_ids).index_by(&:id)
@events.each do |event|
event.network_range = network_ranges[event.network_range_id] if event.network_range_id
end
end
# Debug info
Rails.logger.debug "Events count before pagination: #{@events.count}"
# Load rule associations if needed
rule_ids = @events.map(&:rule_id).compact.uniq
if rule_ids.any?
rules = Rule.where(id: rule_ids).index_by(&:id)
@events.each do |event|
event.rule = rules[event.rule_id] if event.rule_id
end
end
# Paginate
@pagy, @events = pagy(@events, items: 50)
Rails.logger.debug "[DuckDB] Found #{result[:total_count]} total events, showing page #{result[:page]}"
else
# Fallback to PostgreSQL
Rails.logger.warn "[EventsController] DuckDB unavailable, falling back to PostgreSQL"
# Network ranges are now preloaded via includes(:network_range)
# The denormalized network_range_id makes this much faster than IP containment lookups
@events = Event.includes(:network_range, :rule).order(timestamp: :desc)
Rails.logger.debug "Events count after pagination: #{@events.count}"
Rails.logger.debug "Pagy info: #{@pagy.count} total, #{@pagy.pages} pages"
# Apply filters using ActiveRecord scopes
@events = @events.by_ip(params[:ip]) if params[:ip].present?
@events = @events.by_waf_action(params[:waf_action]) if params[:waf_action].present?
@events = @events.by_country(params[:country]) if params[:country].present?
@events = @events.where(rule_id: params[:rule_id]) if params[:rule_id].present?
@events = @events.by_company(params[:company]) if params[:company].present?
@events = @events.by_network_type(params[:network_type]) if params[:network_type].present?
@events = @events.by_asn(params[:asn]) if params[:asn].present?
@events = @events.by_network_cidr(params[:network_cidr]) if params[:network_cidr].present?
@events = @events.exclude_bots if params[:exclude_bots] == "true"
# Paginate
@pagy, @events = pagy(@events, items: 50)
Rails.logger.debug "[PostgreSQL] Events count: #{@pagy.count} total, #{@pagy.pages} pages"
end
end
end