Module: Legion::LLM::Fleet::Dispatcher

Extended by:
Legion::Logging::Helper
Defined in:
lib/legion/llm/fleet/dispatcher.rb

Class Method Summary collapse

Class Method Details

.build_routing_key(provider:, request_type:, model:) ⇒ Object 



55
56
57
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 55

def build_routing_key(provider:, request_type:, model:)
  "llm.request.#{provider}.#{request_type}.#{sanitize_model(model)}"
end

.dispatch(model: nil, messages: nil, request: nil, message_context: {}, routing_key: nil, reply_to: nil, **opts) ⇒ Object

Backwards-compatible shim: supports old (model:, messages:) and new (request:, message_context:) callers



14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 14

def dispatch(model: nil, messages: nil, request: nil, message_context: {}, routing_key: nil, reply_to: nil, **opts)
  return error_result('fleet_unavailable', message_context: message_context) unless fleet_available?

  # Old calling convention: build minimal params from model/messages
  if request.nil? && (model || messages)
    provider = opts[:provider] || 'ollama'
    request_type = opts[:request_type] || 'chat'
    routing_key ||= build_routing_key(provider: provider, request_type: request_type, model: model)
    reply_to ||= ReplyDispatcher.agent_queue_name
    correlation_id = publish_request(
      routing_key: routing_key, reply_to: reply_to,
      provider: provider, model: model, request_type: request_type,
      messages: messages, message_context: message_context, **opts
    )
    timeout = resolve_timeout(request_type: request_type, override: opts[:timeout])
    return wait_for_response(correlation_id, timeout: timeout, message_context: message_context)
  end

  # New calling convention
  request_opts =
    if request.respond_to?(:to_h)
      request.to_h.transform_keys(&:to_sym)
    else
      {}
    end
  request_opts = request_opts.merge(opts)

  provider = request_opts[:provider] || 'ollama'
  request_type = request_opts[:request_type] || 'chat'
  model = request_opts[:model]
  routing_key ||= build_routing_key(provider: provider, request_type: request_type, model: model)
  reply_to ||= ReplyDispatcher.agent_queue_name
  correlation_id = publish_request(
    routing_key: routing_key, reply_to: reply_to,
    provider: provider, model: model, request_type: request_type,
    message_context: message_context, **request_opts.except(:provider, :model, :request_type, :timeout)
  )
  timeout = resolve_timeout(request_type: request_type, override: request_opts[:timeout] || opts[:timeout])
  wait_for_response(correlation_id, timeout: timeout, message_context: message_context)
end

.error_result(reason, message_context: {}) ⇒ Object 



128
129
130
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 128

def error_result(reason, message_context: {})
  { success: false, error: reason, message_context: message_context }
end

.fleet_available? ⇒ Boolean

Returns:

  • (Boolean) 


63
64
65
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 63

def fleet_available?
  transport_ready? && fleet_enabled?
end

.fleet_enabled? ⇒ Boolean

Returns:

  • (Boolean) 


72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 72

def fleet_enabled?
  return true unless defined?(Legion::Settings)

  settings = begin
    Legion::Settings[:llm]
  rescue StandardError => e
    handle_exception(e, level: :debug, operation: 'llm.fleet.dispatcher.fleet_enabled')
    nil
  end
  return true unless settings.is_a?(Hash)

  routing = settings[:routing]
  return true unless routing.is_a?(Hash)

  routing.fetch(:use_fleet, true)
end

.publish_request(**opts) ⇒ Object 



99
100
101
102
103
104
105
106
107
108
109
110
111
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 99

def publish_request(**opts)
  correlation_id = "req_#{SecureRandom.uuid}"
  opts[:fleet_correlation_id] = correlation_id
  log.debug("[llm][fleet][dispatcher] action=publish_request correlation_id=#{correlation_id} routing_key=#{opts[:routing_key]}")

  if defined?(Legion::LLM::Transport::Messages::FleetRequest)
    Legion::LLM::Transport::Messages::FleetRequest.new(**opts).publish
  else
    log.debug('[llm][fleet][dispatcher] action=skip_publish reason=transport_not_loaded')
  end

  correlation_id
end

.resolve_timeout(request_type: :default, override: nil) ⇒ Object 



89
90
91
92
93
94
95
96
97
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 89

def resolve_timeout(request_type: :default, override: nil)
  return override if override

  fleet = Legion::LLM.settings.dig(:routing, :tiers, :fleet) || {}
  fleet.dig(:timeouts, request_type.to_sym) || fleet[:timeout_seconds] || 30
rescue StandardError => e
  handle_exception(e, level: :debug, operation: 'llm.fleet.dispatcher.resolve_timeout')
  30
end

.sanitize_model(model) ⇒ Object 



59
60
61
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 59

def sanitize_model(model)
  model.to_s.gsub(':', '.')
end

.timeout_result(correlation_id, timeout, message_context: {}) ⇒ Object 



123
124
125
126
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 123

def timeout_result(correlation_id, timeout, message_context: {})
  { success: false, error: 'fleet_timeout', correlation_id: correlation_id,
    timeout: timeout, message_context: message_context }
end

.transport_ready? ⇒ Boolean

Returns:

  • (Boolean) 


67
68
69
70
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 67

def transport_ready?
  !!(defined?(Legion::Settings) &&
     Legion::Settings[:transport][:connected] == true)
end

.wait_for_response(correlation_id, timeout:, message_context: {}) ⇒ Object 



113
114
115
116
117
118
119
120
121
 
# File 'lib/legion/llm/fleet/dispatcher.rb', line 113

def wait_for_response(correlation_id, timeout:, message_context: {})
  future = ReplyDispatcher.register(correlation_id)
  result = future.value!(timeout)
  result || timeout_result(correlation_id, timeout, message_context: message_context)
rescue Concurrent::CancelledOperationError
  timeout_result(correlation_id, timeout, message_context: message_context)
ensure
  ReplyDispatcher.deregister(correlation_id)
end