casting issues for nightlies

smartsim/_core/control/job.py

@@ -41,7 +41,7 @@ class Job:
     def __init__(
         self,
         job_name: str,
-        job_id: str,
+        job_id: t.Optional[str],
         entity: t.Union[SmartSimEntity, EntityList],
         launcher: str,
         is_task: bool,
@@ -107,7 +107,7 @@ def record_history(self) -> None:
         """Record the launching history of a job."""
         self.history.record(self.jid, self.status, self.returncode, self.elapsed)
 
-    def reset(self, new_job_name: str, new_job_id: str, is_task: bool) -> None:
+    def reset(self, new_job_name: str, new_job_id: t.Optional[str], is_task: bool) -> None:
         """Reset the job in order to be able to restart it.
 
         :param new_job_name: name of the new job step
@@ -174,12 +174,12 @@ def __init__(self, runs: int = 0) -> None:
         :type runs: int, optional
         """
         self.runs = runs
-        self.jids: t.Dict[int, str] = dict()
+        self.jids: t.Dict[int, t.Optional[str]] = dict()
         self.statuses: t.Dict[int, str] = dict()
         self.returns: t.Dict[int, t.Optional[int]] = dict()
         self.job_times: t.Dict[int, float] = dict()
 
-    def record(self, job_id: str, status: str, returncode: t.Optional[int], job_time: float) -> None:
+    def record(self, job_id: t.Optional[str], status: str, returncode: t.Optional[int], job_time: float) -> None:
         """record the history of a job"""
         self.jids[self.runs] = job_id
         self.statuses[self.runs] = status

smartsim/_core/control/jobmanager.py

@@ -168,7 +168,7 @@ def __call__(self) -> t.Dict[str, Job]:
         all_jobs = {**self.jobs, **self.db_jobs}
         return all_jobs
 
-    def add_job(self, job_name: str, job_id: str, entity: t.Union[SmartSimEntity, EntityList], is_task: bool = True) -> None:
+    def add_job(self, job_name: str, job_id: t.Optional[str], entity: t.Union[SmartSimEntity, EntityList], is_task: bool = True) -> None:
         """Add a job to the job manager which holds specific jobs by type.
 
         :param job_name: name of the job step
@@ -278,7 +278,7 @@ def query_restart(self, entity_name: str) -> bool:
             return True
         return False
 
-    def restart_job(self, job_name: str, job_id: str, entity_name: str, is_task: bool = True) -> None:
+    def restart_job(self, job_name: str, job_id: t.Optional[str], entity_name: str, is_task: bool = True) -> None:
         """Function to reset a job to record history and be
         ready to launch again.
 

smartsim/_core/launcher/cobalt/cobaltLauncher.py

@@ -82,7 +82,7 @@ def supported_rs(self) -> t.Dict[t.Type[SettingsBase], t.Type[Step]]:
         }    
 
 
-    def run(self, step: Step) -> str:
+    def run(self, step: Step) -> t.Optional[str]:
         """Run a job step through Cobalt
 
         :param step: a job step instance
@@ -120,8 +120,6 @@ def run(self, step: Step) -> str:
         # if batch submission did not successfully retrieve job ID
         if not step_id and step.managed:
             step_id = self._get_cobalt_step_id(step)
-        if not step_id:
-            raise ValueError("Unable to get step id for job step")
 
         self.step_mapping.add(step.name, step_id, task_id, step.managed)
         return step_id

smartsim/_core/launcher/launcher.py

@@ -64,7 +64,7 @@ def get_step_nodes(self, step_names: t.List[str]) -> t.List[t.List[str]]:
         raise NotImplementedError
 
     @abc.abstractmethod
-    def run(self, step: Step) -> str:
+    def run(self, step: Step) -> t.Optional[str]:
         raise NotImplementedError
 
     @abc.abstractmethod
@@ -117,7 +117,7 @@ def create_step(self, name: str, cwd: str, step_settings: SettingsBase) -> Step:
     def get_step_nodes(self, step_names: t.List[str]) -> t.List[t.List[str]]:  # pragma: no cover
         raise SSUnsupportedError("Node acquisition not supported for this launcher")
 
-    def run(self, step: Step) -> str:  # pragma: no cover
+    def run(self, step: Step) -> t.Optional[str]:  # pragma: no cover
         raise NotImplementedError
 
     def stop(self, step_name: str) -> StepInfo:  # pragma: no cover

smartsim/_core/launcher/lsf/lsfLauncher.py

@@ -80,7 +80,7 @@ def supported_rs(self) -> t.Dict[t.Type[SettingsBase], t.Type[Step]]:
             RunSettings: LocalStep,
         }
 
-    def run(self, step: Step) -> str:
+    def run(self, step: Step) -> t.Optional[str]:
         """Run a job step through LSF
 
         :param step: a job step instance
@@ -116,8 +116,6 @@ def run(self, step: Step) -> str:
             task_id = self.task_manager.start_task(
                 cmd_list, step.cwd, out=output.fileno(), err=error.fileno()
             )
-        if not step_id:
-            raise ValueError("Unable to get step id for job step")
 
         self.step_mapping.add(step.name, step_id, task_id, step.managed)
         return step_id

smartsim/_core/launcher/pbs/pbsLauncher.py

@@ -76,7 +76,7 @@ def supported_rs(self) -> t.Dict[t.Type[SettingsBase], t.Type[Step]]:
             PalsMpiexecSettings: MpiexecStep,
         }
 
-    def run(self, step: Step) -> str:
+    def run(self, step: Step) -> t.Optional[str]:
         """Run a job step through PBSPro
 
         :param step: a job step instance
@@ -112,9 +112,6 @@ def run(self, step: Step) -> str:
         if not step_id and step.managed:
             step_id = self._get_pbs_step_id(step)
 
-        if not step_id:
-            raise ValueError("Unable to get step id for job step")
-
         self.step_mapping.add(step.name, step_id, task_id, step.managed)
 
         return step_id

smartsim/_core/launcher/slurm/slurmLauncher.py

@@ -113,7 +113,7 @@ def get_step_nodes(self, step_names: t.List[str]) -> t.List[t.List[str]]:
             raise LauncherError("Failed to retrieve nodelist from stat")
         return node_lists
 
-    def run(self, step: Step) -> str:
+    def run(self, step: Step) -> t.Optional[str]:
         """Run a job step through Slurm
 
         :param step: a job step instance
@@ -156,9 +156,6 @@ def run(self, step: Step) -> str:
         if not step_id and step.managed:
             step_id = self._get_slurm_step_id(step)
 
-        if not step_id:
-            raise ValueError("Unable to get step id for job step")
-
         self.step_mapping.add(step.name, step_id, task_id, step.managed)
 
         # give slurm a rest
@@ -243,7 +240,7 @@ def _get_managed_step_update(self, step_ids: t.List[str]) -> t.Optional[t.List[S
             _rc = int(stat_tuple[1]) if stat_tuple[1] else None
             info = SlurmStepInfo(stat_tuple[0], _rc)
 
-            task_id = self.step_mapping.get_task_id(int(step_id))
+            task_id = self.step_mapping.get_task_id(step_id)
             if task_id:
                 # we still check the task manager for jobs that didn't ever
                 # become a fully managed job (e.g. error in slurm arguments)

smartsim/_core/launcher/stepMapping.py

@@ -34,7 +34,7 @@
 class StepMap:
     def __init__(self,
                  step_id: t.Optional[str] = None,
-                 task_id: t.Optional[int] = None,
+                 task_id: t.Optional[str] = None,
                  managed: t.Optional[bool] = None) -> None:
         self.step_id = step_id
         self.task_id = task_id
@@ -60,13 +60,13 @@ def add(
         managed: bool = True,
     ) -> None:
         try:
-            n_task_id = int(task_id) if task_id else None
+            n_task_id = str(task_id) if task_id else None
             self.mapping[step_name] = StepMap(step_id, n_task_id, managed)
         except Exception as e:
             msg = f"Could not add step {step_name} to mapping: {e}"
             logger.exception(msg)
 
-    def get_task_id(self, step_id: int) -> t.Optional[int]:
+    def get_task_id(self, step_id: str) -> t.Optional[str]:
         """Get the task id from the step id"""
         task_id = None
         for stepmap in self.mapping.values():

smartsim/entity/dbnode.py

@@ -176,7 +176,7 @@ def _parse_db_host(self, filepath: t.Optional[str] = None) -> str:
             try:
                 ip = self._parse_ips(filepath, 1)[0]
             # suppress error
-            except FileNotFoundError:
+            except (FileNotFoundError, IndexError):
                 pass
 
             logger.debug("Waiting for Redis output files to populate...")