Class: WorkflowManager::Server
- Inherits:
-
Object
- Object
- WorkflowManager::Server
- Defined in:
- lib/workflow_manager/server.rb
Defined Under Namespace
Classes: Config, KyotoDB, PStoreDB, RedisDB
Constant Summary collapse
- @@config =
nil
Class Method Summary collapse
Instance Method Summary collapse
- #cluster_node_list ⇒ Object
- #cluster_nodes ⇒ Object
- #config ⇒ Object
- #copy_commands(org_dir, dest_parent_dir, now = nil) ⇒ Object
- #default_node ⇒ Object
- #delete_command(target) ⇒ Object
- #finalize_monitoring(current_status, log_file, log_dir) ⇒ Object
- #get_log(job_id, with_err = false) ⇒ Object
- #get_script(job_id) ⇒ Object
- #get_script_path(job_id) ⇒ Object
- #hello ⇒ Object
-
#initialize ⇒ Server
constructor
A new instance of Server.
- #input_dataset_exist?(file_list) ⇒ Boolean
- #input_dataset_file_list(dataset_tsv_path) ⇒ Object
- #input_dataset_tsv_path(script_content) ⇒ Object
- #job_list(with_results = false, project_number = nil, job_ids: nil) ⇒ Object
- #kill_job(job_id) ⇒ Object
- #load_dataset_tree(project_number) ⇒ Object
- #log_puts(str) ⇒ Object
- #recovery_job_checker ⇒ Object
- #save_dataset_tree(project_number, json) ⇒ Object
- #start_monitoring(submit_command, user = 'sushi lover', resubmit = 0, script = '', project_number = 0, sge_options = '', log_dir = '') ⇒ Object
- #start_monitoring2(script_path, script_content, user = 'sushi_lover', project_number = 0, sge_options = '', log_dir = '') ⇒ Object
- #start_monitoring3(script_path, script_content, user = 'sushi_lover', project_number = 0, sge_options = '', log_dir = '', next_dataset_id = '') ⇒ Object
- #status(job_id, new_status = nil) ⇒ Object
- #success_or_fail(job_id, log_file) ⇒ Object (also: #check_status)
- #update_time_status(job_id, current_status, script_name, user, project_number) ⇒ Object
Constructor Details
#initialize ⇒ Server
Returns a new instance of Server.
130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 |
# File 'lib/workflow_manager/server.rb', line 130 def initialize @interval = config.interval @resubmit = config.resubmit extension = case DB_MODE when "PStore" '.pstore' when "KyotoCabinet" '.kch' when "Redis" @redis_conf = config.redis_conf '.rdb' end @db_stat = File.join(config.db_dir, 'statuses'+extension) @db_logs = File.join(config.db_dir, 'logs'+extension) @log_dir = File.(config.log_dir) @db_dir = File.(config.db_dir) FileUtils.mkdir_p @log_dir unless File.exist?(@log_dir) FileUtils.mkdir_p @db_dir unless File.exist?(@db_dir) @statuses = case DB_MODE when "PStore" PStoreDB.new(@db_stat) when "KyotoCabinet" KyotoDB.new(@db_stat) when "Redis" RedisDB.new(0, @redis_conf) end @logs = case DB_MODE when "PStore" PStoreDB.new(@db_logs) when "KyotoCabinet" KyotoDB.new(@db_logs) when "Redis" RedisDB.new(1, @redis_conf) end @jobs = RedisDB.new(2, @redis_conf) @trees = RedisDB.new(4, @redis_conf) @system_log = File.join(@log_dir, "system.log") @mutex = Mutex.new @cluster = config.cluster puts("DB = #{DB_MODE}") if DB_MODE == "Redis" puts("Redis conf = #{config.redis_conf}") puts("Redis port = #{@logs.port}") end puts("Cluster = #{@cluster.name}") log_puts("DB = #{DB_MODE}") log_puts("Cluster = #{@cluster.name}") log_puts("Server starts") log_puts("Recovery check") sleep 2 recovery_job_checker end |
Class Method Details
.config ⇒ Object
64 65 66 |
# File 'lib/workflow_manager/server.rb', line 64 def self.config @@config end |
.config=(config) ⇒ Object
61 62 63 |
# File 'lib/workflow_manager/server.rb', line 61 def self.config=(config) @@config = config end |
.configure {|@@config| ... } ⇒ Object
70 71 72 73 74 75 76 77 78 79 80 81 82 |
# File 'lib/workflow_manager/server.rb', line 70 def self.configure @@config = Config.new # default values @@config.log_dir = LOG_DIR @@config.db_dir = DB_DIR @@config.interval = INTERVAL # interval to check jobs, [s] @@config.resubmit = RESUBMIT # how many times at maximum to resubmit when job fails yield(@@config) if @@config.cluster @@config.cluster.log_dir = File.(@@config.log_dir) end @@config end |
Instance Method Details
#cluster_node_list ⇒ Object
556 557 558 |
# File 'lib/workflow_manager/server.rb', line 556 def cluster_node_list @cluster.node_list end |
#cluster_nodes ⇒ Object
214 215 216 |
# File 'lib/workflow_manager/server.rb', line 214 def cluster_nodes @cluster.cluster_nodes end |
#config ⇒ Object
67 68 69 |
# File 'lib/workflow_manager/server.rb', line 67 def config @@config ||= WorkflowManager.configure{} end |
#copy_commands(org_dir, dest_parent_dir, now = nil) ⇒ Object
204 205 206 |
# File 'lib/workflow_manager/server.rb', line 204 def copy_commands(org_dir, dest_parent_dir, now=nil) @cluster.copy_commands(org_dir, dest_parent_dir, now) end |
#default_node ⇒ Object
217 218 219 |
# File 'lib/workflow_manager/server.rb', line 217 def default_node @cluster.default_node.to_s end |
#delete_command(target) ⇒ Object
211 212 213 |
# File 'lib/workflow_manager/server.rb', line 211 def delete_command(target) @cluster.delete_command(target) end |
#finalize_monitoring(current_status, log_file, log_dir) ⇒ Object
284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 |
# File 'lib/workflow_manager/server.rb', line 284 def finalize_monitoring(current_status, log_file, log_dir) if current_status == 'success' or current_status == 'fail' unless log_dir.empty? copy_commands(log_file, log_dir).each do |command| log_puts(command) system command end err_file = log_file.gsub('_o.log','_e.log') copy_commands(err_file, log_dir).each do |command| log_puts(command) system command end end Thread.current.kill end end |
#get_log(job_id, with_err = false) ⇒ Object
501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 |
# File 'lib/workflow_manager/server.rb', line 501 def get_log(job_id, with_err=false) log_file = nil @logs.transaction do |logs| log_file = logs[job_id.to_s] end log_data = if log_file and File.exist?(log_file) "__STDOUT LOG__\n\n" + File.read(log_file) else 'no log file' end if with_err err_file = log_file.gsub(/_o\.log/,'_e.log') if err_file and File.exist?(err_file) log_data << "\n\n__STDERR LOG__\n\n" log_data << File.read(err_file) end end log_data end |
#get_script(job_id) ⇒ Object
520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 |
# File 'lib/workflow_manager/server.rb', line 520 def get_script(job_id) script_file = nil @logs.transaction do |logs| script_file = logs[job_id.to_s] end if script_file script_file = script_file.gsub(/_o\.log/,'') end script = if script_file and File.exist?(script_file) File.read(script_file) else 'no script file' end script end |
#get_script_path(job_id) ⇒ Object
535 536 537 538 539 540 541 542 543 |
# File 'lib/workflow_manager/server.rb', line 535 def get_script_path(job_id) script_file = nil @logs.transaction do |logs| script_file = logs[job_id.to_s] end script_path = if script_file and File.exist?(script_file) script_file.gsub(/_o\.log/,'') end end |
#hello ⇒ Object
201 202 203 |
# File 'lib/workflow_manager/server.rb', line 201 def hello 'hello hoge hoge bar boo bundle, '+ @cluster.name end |
#input_dataset_exist?(file_list) ⇒ Boolean
252 253 254 255 256 257 258 259 260 261 |
# File 'lib/workflow_manager/server.rb', line 252 def input_dataset_exist?(file_list) flag = true file_list.each do |file| unless File.exist?(file) flag = false break end end flag end |
#input_dataset_file_list(dataset_tsv_path) ⇒ Object
241 242 243 244 245 246 247 248 249 250 251 |
# File 'lib/workflow_manager/server.rb', line 241 def input_dataset_file_list(dataset_tsv_path) file_list = [] CSV.foreach(dataset_tsv_path, :headers=>true, :col_sep=>"\t") do |row| row.each do |header, value| if header =~ /\[File\]/ file_list << value end end end file_list end |
#input_dataset_tsv_path(script_content) ⇒ Object
228 229 230 231 232 233 234 235 236 237 238 239 240 |
# File 'lib/workflow_manager/server.rb', line 228 def input_dataset_tsv_path(script_content) gstore_dir = nil input_dataset_path = nil script_content.split(/\n/).each do |line| if line =~ /GSTORE_DIR=(.+)/ gstore_dir = $1.chomp elsif line =~ /INPUT_DATASET=(.+)/ input_dataset_path = $1.chomp break end end [gstore_dir, input_dataset_path] end |
#job_list(with_results = false, project_number = nil, job_ids: nil) ⇒ Object
472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 |
# File 'lib/workflow_manager/server.rb', line 472 def job_list(with_results=false, project_number=nil, job_ids:nil) s = [] job_idsh = if job_ids Hash[*(job_ids.split(',')).map{|job_id| [job_id, true]}.flatten] end if project_number s_ = {} @jobs.transaction do |jobs| if project_jobs = jobs[project_number] s_ = Hash[*eval(project_jobs)] end end @statuses.transaction do |statuses| s_.each do |job_id, stat| s << [job_id, statuses[job_id]] end end else @statuses.transaction do |statuses| statuses.each do |key, value| s << [key, value] end end end if job_ids s = s.select{|job_id, stat| job_idsh[job_id]} end s.sort_by{|key, value| value.split(',')[2]}.reverse.map{|v| v.join(',')}.join("\n") end |
#kill_job(job_id) ⇒ Object
207 208 209 210 |
# File 'lib/workflow_manager/server.rb', line 207 def kill_job(job_id) status(job_id, 'FAIL') status = `#{@cluster.kill_command(job_id)}` end |
#load_dataset_tree(project_number) ⇒ Object
565 566 567 568 569 570 571 |
# File 'lib/workflow_manager/server.rb', line 565 def load_dataset_tree(project_number) json = nil @trees.transaction do |trees| json = trees[project_number] end json end |
#log_puts(str) ⇒ Object
220 221 222 223 224 225 226 227 |
# File 'lib/workflow_manager/server.rb', line 220 def log_puts(str) time = Time.now.strftime("[%Y.%m.%d %H:%M:%S]") @mutex.synchronize do open(@system_log, "a") do |out| out.print time + " " + str + "\n" end end end |
#recovery_job_checker ⇒ Object
184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 |
# File 'lib/workflow_manager/server.rb', line 184 def recovery_job_checker @logs.transaction do |logs| @statuses.transaction do |statuses| statuses.each do |job_id, status| # puts [job_id, status].join(",") # 120249,RUNNING,QC_ventricles_100k.sh,2021-07-30 09:47:04/2021-07-30 09:47:04,masaomi,1535 stat, script_basename, time, user, project_number, next_dataset_id = status.split(",") if stat == "RUNNING" or stat == "PENDING" log_file = logs[job_id] log_puts("JobID (in recovery check): #{job_id}") puts "JobID (in recovery check): #{job_id}" JobChecker.perform_async(job_id, script_basename, log_file, user, project_number, next_dataset_id) end end end end end |
#save_dataset_tree(project_number, json) ⇒ Object
559 560 561 562 563 564 |
# File 'lib/workflow_manager/server.rb', line 559 def save_dataset_tree(project_number, json) @trees.transaction do |trees| trees[project_number] = json end json end |
#start_monitoring(submit_command, user = 'sushi lover', resubmit = 0, script = '', project_number = 0, sge_options = '', log_dir = '') ⇒ Object
361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 |
# File 'lib/workflow_manager/server.rb', line 361 def start_monitoring(submit_command, user = 'sushi lover', resubmit = 0, script = '', project_number = 0, ='', log_dir = '') log_puts("monitoring: script=" + submit_command + " user=" + user + " resubmit=" + resubmit.to_s + " project=" + project_number.to_s + " sge option=" + + " log dir=" + log_dir.to_s) #warn submit_command # # TODO: analyze arguments # job_id, log_file, command = @cluster.submit_job(submit_command, script, ) log_puts("submit: " + job_id + " " + command) # # monitor worker # if job_id and log_file monitor_worker = Thread.new(job_id, log_file, submit_command, user, resubmit, script, project_number, , log_dir) do |t_job_id, t_log_file, t_submit_command, t_user, t_resubmit, t_script, t_project_number, , t_log_dir| loop do status = success_or_fail(t_job_id, t_log_file) script_name = File.basename(submit_command).split(/-/).first #@statuses.open(@db_stat) @statuses.transaction do |statuses| #start_time = if stat = @statuses[t_job_id] and stat = stat.split(/,/) and time = stat[2] start_time = if stat = statuses[t_job_id] and stat = stat.split(/,/) and time = stat[2] time end time = if start_time if status == 'success' or status == 'fail' start_time + '/' + Time.now.strftime("%Y-%m-%d %H:%M:%S") else start_time end else Time.now.strftime("%Y-%m-%d %H:%M:%S") end #@statuses[t_job_id] = [status, script_name, time, user, project_number].join(',') statuses[t_job_id] = [status, script_name, time, user, project_number].join(',') #@statuses.close end @logs.transaction do |logs| logs[t_job_id] = t_log_file end #warn t_job_id + " " + status if status == 'success' log_puts(status + ": " + t_job_id) unless t_log_dir.empty? copy_commands(t_log_file, t_log_dir).each do |command| log_puts(command) system command end err_file = t_log_file.gsub('_o.log','_e.log') copy_commands(err_file, t_log_dir).each do |command| log_puts(command) system command end end Thread.current.kill elsif status == 'fail' log_puts(status + ": " + t_job_id) # # TODO: re-submit # if t_resubmit < RESUBMIT log_puts("resubmit: " + t_job_id) resubmit_job_id = start_monitoring(t_submit_command, t_user, t_resubmit + 1, t_script, t_project_number, ) script_name = File.basename(submit_command).split(/-/).first #@statuses.open(@db_stat) @statuses.transaction do |statuses| statuses[t_job_id] = ["resubmit: " + resubmit_job_id.to_s, script_name, Time.now.strftime("%Y-%m-%d %H:%M:%S"), t_user, t_project_number].join(',') #@statuses.close end else log_puts("fail: " + t_job_id) end unless t_log_dir.empty? copy_commands(t_log_file, t_log_dir).each do |command| log_puts(command) system command end err_file = t_log_file.gsub('_o.log','_e.log') copy_commands(err_file, t_log_dir).each do |command| log_puts(command) system command end end Thread.current.kill end sleep @interval end end job_id.to_i end end |
#start_monitoring2(script_path, script_content, user = 'sushi_lover', project_number = 0, sge_options = '', log_dir = '') ⇒ Object
311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 |
# File 'lib/workflow_manager/server.rb', line 311 def start_monitoring2(script_path, script_content, user='sushi_lover', project_number=0, ='', log_dir='') # script_path is only used to generate a log file name # It is not used to read the script contents go_submit = false waiting_time = 0 gstore_dir, input_dataset_path = input_dataset_tsv_path(script_content) if gstore_dir and input_dataset_path waiting_max = 60*60*8 # 8h # wait until the files come until waiting_time > waiting_max or File.exist?(input_dataset_path) and file_list = input_dataset_file_list(input_dataset_path) and file_list.map!{|file| File.join(gstore_dir, file)} and go_submit = input_dataset_exist?(file_list) sleep @interval waiting_time += @interval end end job_id, log_file, command = if go_submit @cluster.submit_job(script_path, script_content, ) else raise "stop submitting #{File.basename(script_path)}, since waiting_time #{waiting_time} > #{waiting_max}" end if job_id and log_file # save log_file in logs @logs.transaction do |logs| logs[job_id] = log_file end # job status check until it finishes with success or fail worker = Thread.new(log_dir, script_path, script_content, ) do |log_dir, script_path, script_content, | loop do # check status current_status = check_status(job_id, log_file) # save time and status update_time_status(job_id, current_status, script_path, user, project_number) # finalize (kill current thred) in case of success or fail finalize_monitoring(current_status, log_file, log_dir) # wait sleep @interval end # loop end job_id end end |
#start_monitoring3(script_path, script_content, user = 'sushi_lover', project_number = 0, sge_options = '', log_dir = '', next_dataset_id = '') ⇒ Object
300 301 302 303 304 305 306 307 308 309 310 |
# File 'lib/workflow_manager/server.rb', line 300 def start_monitoring3(script_path, script_content, user='sushi_lover', project_number=0, ='', log_dir='', next_dataset_id='') script_basename = File.basename(script_path) job_id, log_file, command = @cluster.submit_job(script_path, script_content, ) #p command #p log_file #p job_id puts "JobID (in WorkflowManager): #{job_id}" sleep 1 JobChecker.perform_async(job_id, script_basename, log_file, user, project_number, next_dataset_id) job_id end |
#status(job_id, new_status = nil) ⇒ Object
452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 |
# File 'lib/workflow_manager/server.rb', line 452 def status(job_id, new_status=nil) stat = nil #@statuses.open(@db_stat) @statuses.transaction do |statuses| if new_status and stat = statuses[job_id.to_s] status_list = ['CONPLETED', 'RUNNING', 'PENDING', 'FAIL'] if status_list.include?(new_status) items = stat.split(/,/) items.shift items.unshift(new_status) stat = items.join(',') statuses[job_id.to_s] = stat end else stat = statuses[job_id.to_s] end end #@statuses.close stat end |
#success_or_fail(job_id, log_file) ⇒ Object Also known as: check_status
544 545 546 547 548 549 550 551 552 553 554 |
# File 'lib/workflow_manager/server.rb', line 544 def success_or_fail(job_id, log_file) msg = if @cluster.job_running?(job_id) 'running' elsif @cluster.job_ends?(log_file) 'success' elsif @cluster.job_pending?(job_id) 'pending' else 'fail' end end |
#update_time_status(job_id, current_status, script_name, user, project_number) ⇒ Object
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 |
# File 'lib/workflow_manager/server.rb', line 262 def update_time_status(job_id, current_status, script_name, user, project_number) # if the current status changed from last time, then save, otherwise do nothing # once status changes into success or fail, then the thread is expected to be killed in later process @statuses.transaction do |statuses| start_time = nil if stat = statuses[job_id] last_status, script_name, start_time, user, project_number = stat.split(/,/) end time = if start_time if current_status == 'success' or current_status == 'fail' start_time + '/' + Time.now.strftime("%Y-%m-%d %H:%M:%S") elsif current_status != last_status Time.now.strftime("%Y-%m-%d %H:%M:%S") end else Time.now.strftime("%Y-%m-%d %H:%M:%S") end if time statuses[job_id] = [current_status, script_name, time, user, project_number].join(',') end end end |