def _get_slurm_resume_data()

in src/slurm_plugin/instance_manager.py [0:0]

• 26 lines of code
• 4 McCabe index (conditional complexity)

    def _get_slurm_resume_data(self, slurm_resume: Dict[str, any], node_list: List[str]) -> SlurmResumeData:
        """
        Get SlurmResumeData object.

        SlurmResumeData object contains the following:
            * the node list for jobs allocated to single node
            * the node list for jobs allocated to multiple nodes
            * the job list with single node allocation
            * the job list with multi node allocation

        Example of Slurm Resume File (ref. https://slurm.schedmd.com/elastic_computing.html):
        {
            "all_nodes_resume": "cloud[1-3,7-8]",
            "jobs": [
                {
                    "extra": "An arbitrary string from --extra",
                    "features": "c1,c2",
                    "job_id": 140814,
                    "nodes_alloc": "queue1-st-c5xlarge-[4-5]",
                    "nodes_resume": "queue1-st-c5xlarge-[1,3]",
                    "oversubscribe": "OK",
                    "partition": "cloud",
                    "reservation": "resv_1234",
                },
                {
                    "extra": None,
                    "features": "c1,c2",
                    "job_id": 140815,
                    "nodes_alloc": "queue2-st-c5xlarge-[1-2]",
                    "nodes_resume": "queue2-st-c5xlarge-[1-2]",
                    "oversubscribe": "OK",
                    "partition": "cloud",
                    "reservation": None,
                },
                {
                    "extra": None,
                    "features": None,
                    "job_id": 140816,
                    "nodes_alloc": "queue2-st-c5xlarge-[7,8]",
                    "nodes_resume": "queue2-st-c5xlarge-[7,8]",
                    "oversubscribe": "NO",
                    "partition": "cloud_exclusive",
                    "reservation": None,
                },
            ],
        }
        """
        jobs_single_node = []
        jobs_multi_node = []
        single_node = []
        multi_node = []

        slurm_resume_jobs = self._parse_slurm_resume(slurm_resume)

        for job in slurm_resume_jobs:
            if len(job.nodes_resume) == 1:
                jobs_single_node.append(job)
                single_node.extend(job.nodes_resume)
            else:
                jobs_multi_node.append(job)
                multi_node.extend(job.nodes_resume)

        nodes_difference = list(set(node_list) - (set(single_node) | set(multi_node)))

        if nodes_difference:
            logger.warning(
                "Discarding NodeNames because of mismatch in Slurm Resume File Vs Nodes passed to Resume Program: %s",
                ", ".join(nodes_difference),
            )
            self._update_failed_nodes(set(nodes_difference), "InvalidNodenameError")
        return SlurmResumeData(
            single_node=list(dict.fromkeys(single_node)),
            multi_node=list(dict.fromkeys(multi_node)),
            jobs_single_node=jobs_single_node,
            jobs_multi_node=jobs_multi_node,
        )