Slurm APIs

Overview

All of the Slurm commands utilize a collection of Application Programming Interfaces (APIs). User and system applications can directly use these APIs as desired to achieve tighter integration with Slurm. For example, Slurm data structures and error codes can be directly examined rather than executing Slurm commands and parsing their output. This document describes Slurm APIs. You should see the man pages for individual APIs to get more details.

Get Overall Slurm Information

• slurm_api_version — Get Slurm API version number.
• slurm_load_ctl_conf — Load system-wide configuration specifications. Free with slurm_free_ctl_conf to avoid memory leak.
• slurm_print_ctl_conf — Print system-wide configuration specifications.
• slurm_free_ctl_conf — Free storage allocated by slurm_load_ctl_conf.

Get Job Information

• slurm_pid2jobid — For a given process ID on a node get the corresponding Slurm job ID.
• slurm_get_end_time — For a given Slurm job ID get the expected termination time.
• slurm_load_jobs — Load job information. Free with slurm_free_job_info_msg to avoid memory leak.
• slurm_print_job_info_msg — Print information about all jobs.
• slurm_print_job_info — Print information about a specific job.
• slurm_free_job_info_msg — Free storage allocated by slurm_load_jobs.

Get Job Step Information

• slurm_get_job_steps — Load job step information. Free with slurm_free_job_step_info_response_msg to avoid memory leak.
• slurm_print_job_step_info_msg — Print information about all job steps.
• slurm_print_job_step_info — Print information about a specific job step.
• slurm_free_job_step_info_response_msg — Free storage allocated by slurm_get_job_steps.

Get Node Information

• slurm_load_node — Load node information. Free with slurm_free_node_info to avoid memory leak.
• slurm_print_node_info_msg — Print information about all nodes.
• slurm_print_node_table — Print information about a specific node.
• slurm_free_node_info — Free storage allocated by slurm_load_node.

Get Partition Information

• slurm_load_partitions — Load partition (queue) information. Free with slurm_free_partition_info to avoid memory leak.
• slurm_print_partition_info_msg — Print information about all partitions.
• slurm_print_partition_info — Print information about a specific partition.
• slurm_free_partition_info — Free storage allocated by slurm_load_partitions.

Error Handling

• slurm_get_errno — Return the error code set by the last Slurm API function executed.
• slurm_perror — Prin. Slurm error information to standard output.
• slurm_strerror — Return a string describing a specific Slurm error code.

Resource Allocation

• slurm_init_job_desc_msg — Initialize the data structure used in resource allocation requests. You can then just set the fields of particular interest and let the others use default values.
• slurm_job_will_run — Determine if a job would be immediately initiated if submitted now.
• slurm_allocate_resources — Allocate resources for a job. Response message must be freed using slurm_free_resource_allocation_response_msg to avoid a memory leak.
• slurm_free_resource_allocation_response_msg — Frees memory allocated by slurm_allocate_resources.
• slurm_allocate_resources_and_run — Allocate resources for a job and spawn a job step. Response message must be freed using slurm_free_resource_allocation_and_run_response_msg to avoid a memory leak.
• slurm_free_resource_allocation_and_run_response_msg — Frees memory allocated by slurm_allocate_resources_and_run.
• slurm_submit_batch_job — Submit a script for later execution. Response message must be freed using slurm_free_submit_response_response_msg to avoid a memory leak.
• slurm_free_submit_response_response_msg — Frees memory allocated by slurm_submit_batch_job.
• slurm_confirm_allocation — Test if a resource allocation has already been made for a given job id. Response message must be freed using slurm_free_resource_allocation_response_msg to avoid a memory leak. This can be used to confirm that an allocation is still active or for error recovery.

Job Step Creation

Slurm job steps involve numerous interactions with the slurmd daemon. The job step creation is only the first step in the process. We don't advise direct user creation of job steps, but include the information here for completeness.

• slurm_job_step_create — Initiate a job step. Allocated memory must be freed by slurm_free_job_step_create_response_msg to avoid a memory leak.
• slurm_free_job_step_create_response_msg — Free memory allocated by slurm_job_step_create.
• slurm_step_ctx_create — Create job step context. Destroy using slurm_step_ctx_destroy.
• slurm_step_ctx_destroy — Destroy a job step context created by slurm_step_ctx_create.
• slurm_step_ctx_get — Get values from job step context.
• slurm_step_ctx_set — Set values in job step context.
• slurm_jobinfo_ctx_get — Get values from a jobinfo field as returned by slurm_step_ctx_get.
• slurm_spawn — Spawn tasks and establish communications.
• slurm_spawn_kill — Signal spawned tasks.

Job and Job Step Signaling and Cancelling

• slurm_kill_job — Signal or cancel a job.
• slurm_kill_job_step — Signal or cancel a job step.

Job Completion

• slurm_complete_job — Note completion of a job. Releases resource allocation for the job.
• slurm_complete_job_step — Note completion of a job step.

Administrative Functions

Most of these functions can only be executed by user root.

• slurm_reconfigure — Update slurm daemons based upon current slurm.conf configuration file. Use this after updating the configuration file to ensure that it takes effect.
• slurm_shutdown — Terminate slurm daemons.
• slurm_update_job — Update state information associated with a given job.
• slurm_update_node — Update state information associated with a given node. NOTE: Most of a node's characteristics can not be modified.
• slurm_init_part_desc_msg — Initialize a partition update descriptor. Used this to initialize the data structure used in slurm_update_partition.
• slurm_update_partition — Update state information associated with a given partition.
• slurm_delete_partition — Destroy a partition.

Slurm Host List Support

Slurm uses a condensed format to express node names. For example linux[1-3,6] represents linux1, linux2, linux3, and linux6. These functions permit you to translate the Slurm expression into a list of individual node names.

• slurm_hostlist_create — Translate a Slurm node name expression into a record used for parsing. Use slurm_hostlist_destroy to free the allocated storage.
• slurm_hostlist_shift — Get the next node name.
• slurm_hostlist_destroy — Release storage allocated by slurm_hostlist_create.

Last modified 23 November 2019