Class: OodCore::Job::Adapters::Fujitsu_TCS

Inherits:
OodCore::Job::Adapter show all
Defined in:
lib/ood_core/job/adapters/fujitsu_tcs.rb

Overview

An adapter object that describes the communication with a Fujitsu TCS resource manager for job management.

Defined Under Namespace

Classes: Batch

Constant Summary collapse

STATE_MAP =

Mapping of state codes for Fujitsu TCS resource manager

{
  'ACC' => :queued,    # Accepted job submission
  'RJT' => :completed, # Rejected job submission
  'QUE' => :queued,    # Waiting for job execution
  'RNA' => :queued,    # Acquiring resources required for job execution
  'RNP' => :running,   # Executing prologue
  'RUN' => :running,   # Executing job
  'RNE' => :running,   # Executing epilogue
  'RNO' => :running,   # Waiting for completion of job termination processing
  'SPP' => :suspended, # Suspend in progress
  'SPD' => :suspended, # Suspended
  'RSM' => :running,   # Resume in progress
  'EXT' => :completed, # Exited job end execution
  'CCL' => :completed, # Exited job execution by interruption
  'HLD' => :suspended, # In fixed state due to users
  'ERR' => :completed, # In fixed state due to an error
}

Instance Method Summary collapse

Methods inherited from OodCore::Job::Adapter

#accounts, #cluster_info, #info_all_each, #info_where_owner_each, #job_name_illegal_chars, #nodes, #queues, #sanitize_job_name, #supports_job_arrays?

Constructor Details

#initialize(opts = {}) ⇒ Fujitsu_TCS

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Returns a new instance of Fujitsu_TCS.

Parameters:

  • opts (#to_h) (defaults to: {}) —

    the options defining this adapter

Options Hash (opts):

  • :the (Batch) —

    Fujitsu TCS batch object

See Also:



192
193
194
195
196
 
# File 'lib/ood_core/job/adapters/fujitsu_tcs.rb', line 192

def initialize(opts = {})
  o = opts.to_h.symbolize_keys

  @fujitsu_tcs = o.fetch(:fujitsu_tcs) { raise ArgumentError, "No Fujitsu TCS object specified. Missing argument: fujitsu_tcs" }
end

Instance Method Details

#delete(id) ⇒ void

This method returns an undefined value.

Delete the submitted job

Parameters:

  • id (#to_s) —

    the id of the job

Raises:

See Also:



375
376
377
378
379
380
 
# File 'lib/ood_core/job/adapters/fujitsu_tcs.rb', line 375

def delete(id)
  @fujitsu_tcs.delete_job(id.to_s)
rescue Batch::Error => e
  # assume successful job deletion if can't find job id
  raise JobAdapterError, e.message unless /\[ERR\.\] PJM .+ Job .+ does not exist/ =~ e.message
end

#directive_prefix ⇒ Object 



382
383
384
 
# File 'lib/ood_core/job/adapters/fujitsu_tcs.rb', line 382

def directive_prefix
  '#PJM'
end

#hold(id) ⇒ void

This method returns an undefined value.

Put the submitted job on hold

Parameters:

  • id (#to_s) —

    the id of the job

Raises:

See Also:



351
352
353
354
355
356
 
# File 'lib/ood_core/job/adapters/fujitsu_tcs.rb', line 351

def hold(id)
  @fujitsu_tcs.hold_job(id.to_s)
rescue Batch::Error => e
  # assume successful job hold if can't find job id
  raise JobAdapterError, e.message unless /\[ERR\.\] PJM .+ Job .+ does not exist/ =~ e.message
end

#info(id) ⇒ Info

Retrieve job info from the resource manager

Parameters:

  • id (#to_s) —

    the id of the job

Returns:

  • (Info) —

    information describing submitted job

Raises:

See Also:



288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
 
# File 'lib/ood_core/job/adapters/fujitsu_tcs.rb', line 288

def info(id)
  id = id.to_s
  info_ary = @fujitsu_tcs.get_jobs(id: id).map do |v|
    parse_job_info(v)
  end

  # If no job was found we assume that it has completed
  info_ary.empty? ? Info.new(id: id, status: :completed) : info_ary.first # @fujitsu_tcs.get_jobs() must return only one element.
rescue Batch::Error => e
  # set completed status if can't find job id
  if /\[ERR\.\] PJM .+ Job .+ does not exist/ =~ e.message
    Info.new(
      id: id,
      status: :completed
    )
  else
    raise JobAdapterError, e.message
  end
end

#info_all(attrs: nil) ⇒ Array<Info>

Retrieve info for all jobs from the resource manager

Returns:

  • (Array<Info>) —

    information describing submitted jobs

Raises:

See Also:



275
276
277
278
279
280
281
 
# File 'lib/ood_core/job/adapters/fujitsu_tcs.rb', line 275

def info_all(attrs: nil)
  @fujitsu_tcs.get_jobs().map do |v|
    parse_job_info(v)
  end
rescue Batch::Error => e
  raise JobAdapterError, e.message
end

#info_where_owner(owner, attrs: nil) ⇒ Array<Info>

Retrieve info for all jobs for a given owner or owners from the resource manager

Parameters:

  • owner (#to_s, Array<#to_s>) —

    the owner(s) of the jobs

Returns:

  • (Array<Info>) —

    information describing submitted jobs

Raises:



313
314
315
316
317
318
319
320
 
# File 'lib/ood_core/job/adapters/fujitsu_tcs.rb', line 313

def info_where_owner(owner, attrs: nil)
  owner = Array.wrap(owner).map(&:to_s).join('+')
  @fujitsu_tcs.get_jobs(owner: owner).map do |v|
    parse_job_info(v)
  end
rescue Batch::Error => e
  raise JobAdapterError, e.message
end

#release(id) ⇒ void

This method returns an undefined value.

Release the job that is on hold

Parameters:

  • id (#to_s) —

    the id of the job

Raises:

See Also:



363
364
365
366
367
368
 
# File 'lib/ood_core/job/adapters/fujitsu_tcs.rb', line 363

def release(id)
  @fujitsu_tcs.release_job(id.to_s)
rescue Batch::Error => e
  # assume successful job release if can't find job id
  raise JobAdapterError, e.message unless /\[ERR\.\] PJM .+ Job .+ does not exist/ =~ e.message
end

#status(id) ⇒ Status

Retrieve job status from resource manager

Parameters:

  • id (#to_s) —

    the id of the job

Returns:

Raises:

See Also:



327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
 
# File 'lib/ood_core/job/adapters/fujitsu_tcs.rb', line 327

def status(id)
  id = id.to_s
  jobs = @fujitsu_tcs.get_jobs(id: id)

  if job = jobs.detect { |j| j[:JOB_ID] == id }
    Status.new(state: get_state(job[:ST]))
  else
    # set completed status if can't find job id
    Status.new(state: :completed)
  end
rescue Batch::Error => e
  # set completed status if can't find job id
  if /\[ERR\.\] PJM .+ Job .+ does not exist/ =~ e.message
    Status.new(state: :completed)
  else
    raise JobAdapterError, e.message
  end
end

#submit(script, after: [], afterok: [], afternotok: [], afterany: []) ⇒ String

Submit a job with the attributes defined in the job template instance

Parameters:

  • script (Script) —

    script object that describes the script and attributes for the submitted job

  • after (#to_s, Array<#to_s>) (defaults to: []) —

    this job may be scheduled for execution at any point after dependent jobs have started execution

  • afterok (#to_s, Array<#to_s>) (defaults to: []) —

    this job may be scheduled for execution only after dependent jobs have terminated with no errors

  • afternotok (#to_s, Array<#to_s>) (defaults to: []) —

    this job may be scheduled for execution only after dependent jobs have terminated with errors

  • afterany (#to_s, Array<#to_s>) (defaults to: []) —

    this job may be scheduled for execution after dependent jobs have terminated

Returns:

  • (String) —

    the job id returned after successfully submitting a job

Raises:

See Also:



213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
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
259
260
261
262
263
264
265
266
267
268
269
 
# File 'lib/ood_core/job/adapters/fujitsu_tcs.rb', line 213

def submit(script, after: [], afterok: [], afternotok: [], afterany: [])
  #after      = Array(after).map(&:to_s)
  #afterok    = Array(afterok).map(&:to_s)
  #afternotok = Array(afternotok).map(&:to_s)
  #afterany   = Array(afterany).map(&:to_s)
  if !after.empty? || !afterok.empty? || !afternotok.empty? || !afterany.empty?
    raise JobAdapterError, "Dependency between jobs has not implemented yet."
  end
  
  # Set pjsub options
  args = []
  args.concat (script.rerunnable ? ["--restart"] : ["--norestart"]) unless script.rerunnable.nil?
  args.concat ["--mail-list", script.email.join(",")] unless script.email.nil?
  if script.email_on_started && script.email_on_terminated
    args.concat ["-m", "b,e"]
  elsif script.email_on_started
    args.concat ["-m", "b"]
  elsif script.email_on_terminated
    args.concat ["-m", "e"]
  end

  args.concat ["-N", script.job_name] unless script.job_name.nil?
  args.concat ["-o", script.output_path] unless script.output_path.nil?
  args.concat ['--mpi', "proc=#{script.cores}"] unless script.cores.nil?
  if script.error_path.nil?
    args.concat ["-j"]
  else
    args.concat ["-e", script.error_path]
  end
  args.concat ["-L", "rscgrp=" + script.queue_name] unless script.queue_name.nil?
  args.concat ["-p", script.priority] unless script.priority.nil?
  
  # start_time: <%= Time.local(2023,11,22,13,4).to_i %> in form.yml.erb
  args.concat ["--at", script.start_time.localtime.strftime("%C%y%m%d%H%M")] unless script.start_time.nil?
  args.concat ["-L", "elapse=" + seconds_to_duration(script.wall_time)] unless script.wall_time.nil?
  args.concat ["--bulk", "--sparam", script.job_array_request] unless script.job_array_request.nil?

  # Set environment variables
  envvars = script.job_environment.to_h
  args.concat ["-x", envvars.map{|k,v| "#{k}=#{v}"}.join(",")] unless envvars.empty?
  args.concat ["-X"] if script.copy_environment?

  # Set native options
  args.concat script.native if script.native
  
  # Set content
  content = if script.shell_path.nil?
              script.content
            else
              "#!#{script.shell_path}\n#{script.content}"
            end
  
  # Submit job
  @fujitsu_tcs.submit_string(content, args: args)
rescue Batch::Error => e
  raise JobAdapterError, e.message
end