Module: NewRelic::Agent::Agent::InstanceMethods
- Includes:
- Connect, Start, StartWorkerThread
- Included in:
- NewRelic::Agent::Agent
- Defined in:
- lib/new_relic/agent/agent.rb
Overview
Holds all the methods defined on NewRelic::Agent::Agent instances
Defined Under Namespace
Modules: Connect, Start, StartWorkerThread
Instance Attribute Summary collapse
-
#beacon_configuration ⇒ Object
readonly
a configuration for the Real User Monitoring system - handles things like static setup of the header for inclusion into pages.
-
#cross_process_encoding_bytes ⇒ Object
readonly
Returns the value of attribute cross_process_encoding_bytes.
-
#cross_process_id ⇒ Object
readonly
cross process id’s and encoding.
-
#error_collector ⇒ Object
readonly
error collector is a simple collection of recorded errors.
-
#events ⇒ Object
readonly
Global events dispatcher.
-
#metric_ids ⇒ Object
readonly
a cached set of metric_ids to save the collector some time - it returns a metric id for every metric name we send it, and in the future we transmit using the metric id only.
-
#obfuscator ⇒ Object
readonly
holds a proc that is used to obfuscate sql statements.
-
#record_sql ⇒ Object
readonly
whether we should record raw, obfuscated, or no sql.
-
#service ⇒ Object
service for communicating with collector.
-
#sql_sampler ⇒ Object
readonly
Returns the value of attribute sql_sampler.
-
#stats_engine ⇒ Object
readonly
the statistics engine that holds all the timeslice data.
-
#thread_profiler ⇒ Object
readonly
begins a thread profile session when instructed by agent commands.
-
#transaction_sampler ⇒ Object
readonly
the transaction sampler that handles recording transactions.
-
#url_rules ⇒ Object
readonly
in theory a set of rules applied by the agent to the output of its metrics.
Attributes included from Connect
Instance Method Summary collapse
-
#after_fork(options = {}) ⇒ Object
This method should be called in a forked process after a fork.
-
#end_transaction ⇒ Object
Tells the statistics engine we are ending a transaction.
- #forked? ⇒ Boolean
-
#merge_data_from(data) ⇒ Object
Accepts data as provided by the serialize method and merges it into our current collection of data to send.
-
#pop_trace_execution_flag ⇒ Object
Pop the current trace execution status.
-
#push_trace_execution_flag(should_trace = false) ⇒ Object
Push flag indicating whether we should be tracing in this thread.
-
#record_transaction(duration_seconds, options = {}) ⇒ Object
fakes out a transaction that did not happen in this process by creating apdex, summary metrics, and recording statistics for the transaction.
-
#reset_stats ⇒ Object
Clear out the metric data, errors, and transaction traces, making sure the agent is in a fresh state.
-
#serialize ⇒ Object
Serialize all the important data that the agent might want to send to the server.
-
#set_record_sql(should_record) ⇒ Object
Sets a thread local variable as to whether we should or should not record sql in the current thread.
-
#set_record_tt(should_record) ⇒ Object
Sets a thread local variable as to whether we should or should not record transaction traces in the current thread.
-
#shutdown(options = {}) ⇒ Object
Attempt a graceful shutdown of the agent, running the worker loop if it exists and is running.
-
#start ⇒ Object
Logs a bunch of data and starts the agent, if needed.
-
#start_transaction ⇒ Object
Tells the statistics engine we are starting a new transaction.
-
#started? ⇒ Boolean
True if we have initialized and completed ‘start’.
-
#unsent_errors_size ⇒ Object
Returns the length of the unsent errors array, if it exists, otherwise nil.
-
#unsent_timeslice_data ⇒ Object
Initializes the unsent timeslice data hash, if needed, and returns the number of keys it contains.
-
#unsent_traces_size ⇒ Object
Returns the length of the traces array, if it exists, otherwise nil.
Methods included from Connect
#apdex_f, #connect_retry_period, #connect_settings, #connect_to_server, #connected?, #disconnect, #disconnected?, #environment_for_connect, #finish_setup, #handle_license_error, #handle_unrecoverable_agent_error, #log_collector_messages, #log_connection!, #log_error, #note_connect_failure, #query_server_for_configuration, #should_connect?
Methods included from StartWorkerThread
#catch_errors, #create_and_run_worker_loop, #deferred_work!, #handle_force_disconnect, #handle_force_restart, #handle_other_error, #handle_server_connection_problem, #log_worker_loop_start
Methods included from Start
#already_started?, #check_config_and_start_agent, #connect_in_foreground, #correct_license_length, #defer_for_resque?, #disabled?, #has_correct_license_key?, #has_license_key?, #install_exit_handler, #log_app_names, #log_dispatcher, #log_environment, #log_if, #log_startup, #log_unless, #log_version_and_pid, #monitoring?, #using_forking_dispatcher?, #using_sinatra?, #weird_ruby?
Instance Attribute Details
#beacon_configuration ⇒ Object (readonly)
a configuration for the Real User Monitoring system - handles things like static setup of the header for inclusion into pages
95 96 97 |
# File 'lib/new_relic/agent/agent.rb', line 95 def beacon_configuration @beacon_configuration end |
#cross_process_encoding_bytes ⇒ Object (readonly)
Returns the value of attribute cross_process_encoding_bytes.
98 99 100 |
# File 'lib/new_relic/agent/agent.rb', line 98 def cross_process_encoding_bytes @cross_process_encoding_bytes end |
#cross_process_id ⇒ Object (readonly)
cross process id’s and encoding
97 98 99 |
# File 'lib/new_relic/agent/agent.rb', line 97 def cross_process_id @cross_process_id end |
#error_collector ⇒ Object (readonly)
error collector is a simple collection of recorded errors
82 83 84 |
# File 'lib/new_relic/agent/agent.rb', line 82 def error_collector @error_collector end |
#events ⇒ Object (readonly)
Global events dispatcher. This will provides our primary mechanism for agent-wide events, such as finishing configuration, error notification and request before/after from Rack.
104 105 106 |
# File 'lib/new_relic/agent/agent.rb', line 104 def events @events end |
#metric_ids ⇒ Object (readonly)
a cached set of metric_ids to save the collector some time - it returns a metric id for every metric name we send it, and in the future we transmit using the metric id only
88 89 90 |
# File 'lib/new_relic/agent/agent.rb', line 88 def metric_ids @metric_ids end |
#obfuscator ⇒ Object (readonly)
holds a proc that is used to obfuscate sql statements
73 74 75 |
# File 'lib/new_relic/agent/agent.rb', line 73 def obfuscator @obfuscator end |
#record_sql ⇒ Object (readonly)
whether we should record raw, obfuscated, or no sql
84 85 86 |
# File 'lib/new_relic/agent/agent.rb', line 84 def record_sql @record_sql end |
#service ⇒ Object
service for communicating with collector
100 101 102 |
# File 'lib/new_relic/agent/agent.rb', line 100 def service @service end |
#sql_sampler ⇒ Object (readonly)
Returns the value of attribute sql_sampler.
78 79 80 |
# File 'lib/new_relic/agent/agent.rb', line 78 def sql_sampler @sql_sampler end |
#stats_engine ⇒ Object (readonly)
the statistics engine that holds all the timeslice data
75 76 77 |
# File 'lib/new_relic/agent/agent.rb', line 75 def stats_engine @stats_engine end |
#thread_profiler ⇒ Object (readonly)
begins a thread profile session when instructed by agent commands
80 81 82 |
# File 'lib/new_relic/agent/agent.rb', line 80 def thread_profiler @thread_profiler end |
#transaction_sampler ⇒ Object (readonly)
the transaction sampler that handles recording transactions
77 78 79 |
# File 'lib/new_relic/agent/agent.rb', line 77 def transaction_sampler @transaction_sampler end |
#url_rules ⇒ Object (readonly)
in theory a set of rules applied by the agent to the output of its metrics. Currently unimplemented
91 92 93 |
# File 'lib/new_relic/agent/agent.rb', line 91 def url_rules @url_rules end |
Instance Method Details
#after_fork(options = {}) ⇒ Object
This method should be called in a forked process after a fork. It assumes the parent process initialized the agent, but does not assume the agent started.
The call is idempotent, but not re-entrant.
-
It clears any metrics carried over from the parent process
-
Restarts the sampler thread if necessary
-
Initiates a new agent run and worker loop unless that was done in the parent process and
:force_reconnect
is not true
Options:
-
:force_reconnect => true
to force the spawned process to establish a new connection, such as when forking a long running process. The default is false–it will only connect to the server if the parent had not connected. -
:keep_retrying => false
if we try to initiate a new connection, this tells me to only try it once so this method returns quickly if there is some kind of latency with the server.
182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 |
# File 'lib/new_relic/agent/agent.rb', line 182 def after_fork(={}) @forked = true Agent.config.apply_config(NewRelic::Agent::Configuration::ManualSource.new(), 1) if channel_id = [:report_to_channel] @service = NewRelic::Agent::PipeService.new(channel_id) if connected? @connected_pid = $$ @metric_ids = {} else ::NewRelic::Agent.logger.debug("Child process #{$$} not reporting to non-connected parent.") @service.shutdown(Time.now) disconnect end end return if !Agent.config[:agent_enabled] || !Agent.config[:monitor_mode] || disconnected? || @worker_thread && @worker_thread.alive? ::NewRelic::Agent.logger.debug "Starting the worker thread in #{$$} after forking." # Clear out stats that are left over from parent process reset_stats # Don't ever check to see if this is a spawner. If we're in a forked process # I'm pretty sure we're not also forking new instances. start_worker_thread() @stats_engine.start_sampler_thread end |
#end_transaction ⇒ Object
Tells the statistics engine we are ending a transaction
266 267 268 |
# File 'lib/new_relic/agent/agent.rb', line 266 def end_transaction @stats_engine.end_transaction end |
#forked? ⇒ Boolean
214 215 216 |
# File 'lib/new_relic/agent/agent.rb', line 214 def forked? @forked end |
#merge_data_from(data) ⇒ Object
Accepts data as provided by the serialize method and merges it into our current collection of data to send. Can be dangerous if we re-merge the same data more than once - it will be sent multiple times.
796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 |
# File 'lib/new_relic/agent/agent.rb', line 796 def merge_data_from(data) metrics, transaction_traces, errors = data @stats_engine.merge_data(metrics) if metrics if transaction_traces && transaction_traces.respond_to?(:any?) && transaction_traces.any? if @traces @traces += transaction_traces else @traces = transaction_traces end end if errors && errors.respond_to?(:each) errors.each do |err| @error_collector.add_to_error_queue(err) end end end |
#pop_trace_execution_flag ⇒ Object
Pop the current trace execution status. Restore trace execution status to what it was before we pushed the current flag.
303 304 305 |
# File 'lib/new_relic/agent/agent.rb', line 303 def pop_trace_execution_flag Thread.current[:newrelic_untraced].pop if Thread.current[:newrelic_untraced] end |
#push_trace_execution_flag(should_trace = false) ⇒ Object
Push flag indicating whether we should be tracing in this thread. This uses a stack which allows us to disable tracing children of a transaction without affecting the tracing of the whole transaction
292 293 294 295 296 297 298 299 |
# File 'lib/new_relic/agent/agent.rb', line 292 def push_trace_execution_flag(should_trace=false) value = Thread.current[:newrelic_untraced] if (value.nil?) Thread.current[:newrelic_untraced] = [] end Thread.current[:newrelic_untraced] << should_trace end |
#record_transaction(duration_seconds, options = {}) ⇒ Object
fakes out a transaction that did not happen in this process by creating apdex, summary metrics, and recording statistics for the transaction
This method is deprecated - it may be removed in future versions of the agent
132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 |
# File 'lib/new_relic/agent/agent.rb', line 132 def record_transaction(duration_seconds, ={}) is_error = ['is_error'] || ['error_message'] || ['exception'] metric = ['metric'] metric ||= ['uri'] # normalize this with url rules raise "metric or uri arguments required" unless metric metric_info = NewRelic::MetricParser::MetricParser.for_metric_named(metric) if metric_info.is_web_transaction? NewRelic::Agent::Instrumentation::MetricFrame.record_apdex(metric_info, duration_seconds, duration_seconds, is_error) end metrics = metric_info.summary_metrics metrics << metric metrics.each do |name| stats = stats_engine.get_stats_no_scope(name) stats.record_data_point(duration_seconds) end if is_error if ['exception'] e = ['exception'] elsif ['error_message'] e = StandardError.new ['error_message'] else e = StandardError.new 'Unknown Error' end error_collector.notice_error e, :uri => ['uri'], :metric => metric end # busy time ? end |
#reset_stats ⇒ Object
Clear out the metric data, errors, and transaction traces, making sure the agent is in a fresh state
499 500 501 502 503 504 505 506 |
# File 'lib/new_relic/agent/agent.rb', line 499 def reset_stats @stats_engine.reset_stats @unsent_errors = [] @traces = nil @unsent_timeslice_data = {} @last_harvest_time = Time.now @launch_time = Time.now end |
#serialize ⇒ Object
Serialize all the important data that the agent might want to send to the server. We could be sending this to file ( common in short-running background transactions ) or alternately we could serialize via a pipe or socket to a local aggregation device
781 782 783 784 785 786 787 788 789 |
# File 'lib/new_relic/agent/agent.rb', line 781 def serialize accumulator = [] accumulator[1] = harvest_transaction_traces if @transaction_sampler accumulator[2] = harvest_errors if @error_collector accumulator[0] = harvest_timeslice_data reset_stats @metric_ids = {} accumulator end |
#set_record_sql(should_record) ⇒ Object
Sets a thread local variable as to whether we should or should not record sql in the current thread. Returns the previous value, if there is one
273 274 275 276 277 |
# File 'lib/new_relic/agent/agent.rb', line 273 def set_record_sql(should_record) prev = Thread::current[:record_sql] Thread::current[:record_sql] = should_record prev.nil? || prev end |
#set_record_tt(should_record) ⇒ Object
Sets a thread local variable as to whether we should or should not record transaction traces in the current thread. Returns the previous value, if there is one
282 283 284 285 286 |
# File 'lib/new_relic/agent/agent.rb', line 282 def set_record_tt(should_record) prev = Thread::current[:record_tt] Thread::current[:record_tt] = should_record prev.nil? || prev end |
#shutdown(options = {}) ⇒ Object
Attempt a graceful shutdown of the agent, running the worker loop if it exists and is running.
Options: :force_send => (true/false) # force the agent to send data
228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 |
# File 'lib/new_relic/agent/agent.rb', line 228 def shutdown(={}) run_loop_before_exit = Agent.config[:force_send] return if not started? if @worker_loop @worker_loop.run_task if run_loop_before_exit @worker_loop.stop end ::NewRelic::Agent.logger.info "Starting Agent shutdown" # if litespeed, then ignore all future SIGUSR1 - it's # litespeed trying to shut us down if Agent.config[:dispatcher] == :litespeed Signal.trap("SIGUSR1", "IGNORE") Signal.trap("SIGTERM", "IGNORE") end begin NewRelic::Agent.disable_all_tracing do graceful_disconnect end rescue => e ::NewRelic::Agent.logger.error e end NewRelic::Agent.config.remove_config do |config| config.class == NewRelic::Agent::Configuration::ManualSource || config.class == NewRelic::Agent::Configuration::ServerSource end @started = nil Control.reset end |
#start ⇒ Object
Logs a bunch of data and starts the agent, if needed
482 483 484 485 486 487 488 489 490 491 492 493 494 495 |
# File 'lib/new_relic/agent/agent.rb', line 482 def start return if already_started? || disabled? if defer_for_resque? ::NewRelic::Agent.logger.debug "Deferring startup for Resque in case it daemonizes" return end @started = true @local_host = determine_host log_startup check_config_and_start_agent log_version_and_pid end |
#start_transaction ⇒ Object
Tells the statistics engine we are starting a new transaction
261 262 263 |
# File 'lib/new_relic/agent/agent.rb', line 261 def start_transaction @stats_engine.start_transaction end |
#started? ⇒ Boolean
True if we have initialized and completed ‘start’
219 220 221 |
# File 'lib/new_relic/agent/agent.rb', line 219 def started? @started end |
#unsent_errors_size ⇒ Object
Returns the length of the unsent errors array, if it exists, otherwise nil
109 110 111 |
# File 'lib/new_relic/agent/agent.rb', line 109 def unsent_errors_size @unsent_errors.length if @unsent_errors end |
#unsent_timeslice_data ⇒ Object
Initializes the unsent timeslice data hash, if needed, and returns the number of keys it contains
121 122 123 124 |
# File 'lib/new_relic/agent/agent.rb', line 121 def unsent_timeslice_data @unsent_timeslice_data ||= {} @unsent_timeslice_data.keys.length end |
#unsent_traces_size ⇒ Object
Returns the length of the traces array, if it exists, otherwise nil
115 116 117 |
# File 'lib/new_relic/agent/agent.rb', line 115 def unsent_traces_size @traces.length if @traces end |