crew_controller_sge: *[Maturing]* Create a controller with a Sun Grid Engine (SGE)...

In crew.cluster: Crew Launcher Plugins for Traditional High-Performance Computing Clusters

Create a controller with a Sun Grid Engine (SGE) launcher.

Description

Create an R6 object to submit tasks and launch workers on Sun Grid Engine (SGE) workers.

Usage

crew_controller_sge(
  name = NULL,
  workers = 1L,
  host = NULL,
  port = NULL,
  tls = crew::crew_tls(mode = "automatic"),
  tls_enable = NULL,
  tls_config = NULL,
  serialization = NULL,
  seconds_interval = 0.25,
  seconds_timeout = 60,
  seconds_launch = 86400,
  seconds_idle = 300,
  seconds_wall = Inf,
  seconds_exit = NULL,
  retry_tasks = NULL,
  tasks_max = Inf,
  tasks_timers = 0L,
  reset_globals = TRUE,
  reset_packages = FALSE,
  reset_options = FALSE,
  garbage_collection = FALSE,
  crashes_error = NULL,
  r_arguments = c("--no-save", "--no-restore"),
  crashes_max = 5L,
  backup = NULL,
  options_metrics = crew::crew_options_metrics(),
  options_cluster = crew.cluster::crew_options_sge(),
  verbose = NULL,
  command_submit = NULL,
  command_terminate = NULL,
  command_delete = NULL,
  script_directory = NULL,
  script_lines = NULL,
  sge_cwd = NULL,
  sge_envvars = NULL,
  sge_log_output = NULL,
  sge_log_error = NULL,
  sge_log_join = NULL,
  sge_memory_gigabytes_limit = NULL,
  sge_memory_gigabytes_required = NULL,
  sge_cores = NULL,
  sge_gpu = NULL
)

Arguments

name	Character string, name of the launcher. If the name is NULL, then a name is automatically generated when the launcher starts.
workers	Maximum number of workers to run concurrently when auto-scaling, excluding task retries and manual calls to launch(). Special workers allocated for task retries do not count towards this limit, so the number of workers running at a given time may exceed this maximum. A smaller number of workers may run if the number of executing tasks is smaller than the supplied value of the workers argument.
host	IP address of the mirai client to send and receive tasks. If NULL, the host defaults to the local IP address.
port	TCP port to listen for the workers. If NULL, then an available ephemeral port is automatically chosen. Controllers running simultaneously on the same computer (as in a controller group) must not share the same TCP port.
tls	A TLS configuration object from crew_tls().
tls_enable	Deprecated on 2023-09-15 in version 0.4.1. Use argument tls instead.
tls_config	Deprecated on 2023-09-15 in version 0.4.1. Use argument tls instead.
serialization	Either NULL (default) or an object produced by mirai::serial_config() to control the serialization of data sent to workers. This can help with either more efficient data transfers or to preserve attributes of otherwise non-exportable objects (such as torch tensors or arrow tables). See ?mirai::serial_config for details.
seconds_interval	Number of seconds between polling intervals waiting for certain internal synchronous operations to complete. In certain cases, exponential backoff is used with this argument passed to seconds_max in a crew_throttle() object.
seconds_timeout	Number of seconds until timing out while waiting for certain synchronous operations to complete, such as checking mirai::status().
seconds_launch	Seconds of startup time to allow. A worker is unconditionally assumed to be alive from the moment of its launch until seconds_launch seconds later. After seconds_launch seconds, the worker is only considered alive if it is actively connected to its assign websocket.
seconds_idle	Maximum number of seconds that a worker can idle since the completion of the last task. If exceeded, the worker exits. But the timer does not launch until tasks_timers tasks have completed. See the idletime argument of mirai::daemon(). crew does not excel with perfectly transient workers because it does not micromanage the assignment of tasks to workers, so please allow enough idle time for a new worker to be delegated a new task.
seconds_wall	Soft wall time in seconds. The timer does not launch until tasks_timers tasks have completed. See the walltime argument of mirai::daemon().
seconds_exit	Deprecated on 2023-09-21 in version 0.1.2.9000. No longer necessary.
retry_tasks	Deprecated on 2025-01-13 (crew version 0.10.2.9002).
tasks_max	Maximum number of tasks that a worker will do before exiting. See the maxtasks argument of mirai::daemon(). crew does not excel with perfectly transient workers because it does not micromanage the assignment of tasks to workers, it is recommended to set tasks_max to a value greater than 1.
tasks_timers	Number of tasks to do before activating the timers for seconds_idle and seconds_wall. See the timerstart argument of mirai::daemon().
reset_globals	TRUE to reset global environment variables between tasks, FALSE to leave them alone.
reset_packages	TRUE to unload any packages loaded during a task (runs between each task), FALSE to leave packages alone.
reset_options	TRUE to reset global options to their original state between each task, FALSE otherwise. It is recommended to only set reset_options = TRUE if reset_packages is also TRUE because packages sometimes rely on options they set at loading time.
garbage_collection	TRUE to run garbage collection between tasks, FALSE to skip.
crashes_error	Deprecated on 2025-01-13 (crew version 0.10.2.9002).
r_arguments	Optional character vector of command line arguments to pass to Rscript (non-Windows) or Rscript.exe (Windows) when starting a worker. Example: r_arguments = c("--vanilla", "--max-connections=32").
crashes_max	In rare cases, a worker may exit unexpectedly before it completes its current task. If this happens, pop() returns a status of "crash" instead of "error" for the task. The controller does not automatically retry the task, but you can retry it manually by calling push() again and using the same task name as before. (However, targets pipelines running crew do automatically retry tasks whose workers crashed.) crashes_max is a non-negative integer, and it sets the maximum number of allowable consecutive crashes for a given task. If a task's worker crashes more than crashes_max times in a row, then pop() throws an error when it tries to return the results of the task.
backup	An optional crew controller object, or NULL to omit. If supplied, the backup controller runs any pushed tasks that have already reached crashes_max consecutive crashes. Using backup, you can create a chain of controllers with different levels of resources (such as worker memory and CPUs) so that a task that fails on one controller can retry using incrementally more powerful workers. All controllers in a backup chain should be part of the same controller group (see crew_controller_group()) so you can call the group-level pop() and collect() methods to make sure you get results regardless of which controller actually ended up running the task. Limitations of backup: * crashes_max needs to be positive in order for backup to be used. Otherwise, every task would always skip the current controller and go to backup. * backup cannot be a controller group. It must be an ordinary controller.
options_metrics	Either NULL to opt out of resource metric logging for workers, or an object from crew_options_metrics() to enable and configure resource metric logging for workers. For resource logging to run, the autometric R package version 0.1.0 or higher must be installed.
options_cluster	An options list from crew_options_sge() with cluster-specific configuration options.
verbose	Deprecated. Use options_cluster instead.
command_submit	Deprecated. Use options_cluster instead.
command_terminate	Deprecated. Use options_cluster instead.
command_delete	Deprecated on 2024-01-08 (version 0.1.4.9001). Use command_terminate instead.
script_directory	Deprecated. Use options_cluster instead.
script_lines	Deprecated. Use options_cluster instead.
sge_cwd	Deprecated. Use options_cluster instead.
sge_envvars	Deprecated. Use options_cluster instead.
sge_log_output	Deprecated. Use options_cluster instead.
sge_log_error	Deprecated. Use options_cluster instead.
sge_log_join	Deprecated. Use options_cluster instead.
sge_memory_gigabytes_limit	Deprecated. Use options_cluster instead.
sge_memory_gigabytes_required	Deprecated. Use options_cluster instead.
sge_cores	Deprecated. Use options_cluster instead.
sge_gpu	Deprecated. Use options_cluster instead.

Attribution

The template files at https://github.com/mschubert/clustermq/tree/master/inst informed the development of the crew launcher plugins in crew.cluster, and we would like to thank Michael Schubert for developing clustermq and releasing it under the permissive Apache License 2.0. See the NOTICE and README.md files in the crew.cluster source code for additional attribution.

See Also

Other sge: crew_class_launcher_sge, crew_class_monitor_sge, crew_launcher_sge(), crew_monitor_sge(), crew_options_sge()

Examples

if (identical(Sys.getenv("CREW_EXAMPLES"), "true")) {
controller <- crew_controller_sge()
controller$start()
controller$push(name = "task", command = sqrt(4))
controller$wait()
controller$pop()$result
controller$terminate()
}

crew.cluster documentation built on April 15, 2025, 1:34 a.m.