Merge pull request #192 from eth-cscs/Post-job-hot-fix

Post job hot fix

Author: theely

src/firecrest/compute/models.py

@@ -65,7 +65,7 @@ class GetJobMetadataResponse(CamelModel):
 
 
 class PostJobSubmissionResponse(CamelModel):
-    job_id: Optional[int] = Field(None, nullable=True)
+    job_id: Optional[str] = Field(None, nullable=True)
 
 
 class PostJobAttachRequest(CamelModel):

src/firecrest/filesystem/transfer/models.py

@@ -81,7 +81,7 @@ class TransferJobLogs(CamelModel):
 
 
 class TransferJob(CamelModel):
-    job_id: int
+    job_id: str
     system: str
     working_directory: str
     logs: TransferJobLogs

src/lib/datatransfers/datatransfer_base.py

@@ -18,7 +18,7 @@ class TransferJobLogs(CamelModel):
 
 
 class TransferJob(CamelModel):
-    job_id: int
+    job_id: str
     system: str
     working_directory: str
     logs: TransferJobLogs

src/lib/scheduler_clients/scheduler_base_client.py

@@ -25,7 +25,7 @@ async def submit_job(
         job_description: JobDescriptionModel,
         username: str,
         jwt_token: str,
-    ) -> int | None:
+    ) -> str | None:
         pass
 
     @abstractmethod
@@ -35,17 +35,13 @@ async def attach_command(
         job_id: str,
         username: str,
         jwt_token: str,
-    ) -> int | None:
+    ) -> None:
         pass
 
     @abstractmethod
     # Note: returns multiple jobs to deal with job_id duplicates (see Slurm doc)
     async def get_job(
-        self,
-        job_id: str,
-        username: str,
-        jwt_token: str,
-        allusers: bool = True
+        self, job_id: str, username: str, jwt_token: str, allusers: bool = True
     ) -> List[JobModel]:
         pass
 
@@ -58,10 +54,7 @@ async def get_job_metadata(
 
     @abstractmethod
     async def get_jobs(
-        self,
-        username: str,
-        jwt_token: str,
-        allusers: bool = False
+        self, username: str, jwt_token: str, allusers: bool = False
     ) -> List[JobModel] | None:
         pass
 

src/lib/scheduler_clients/slurm/slurm_base_client.py

@@ -25,7 +25,7 @@ async def submit_job(
         job_description: SlurmJobDescription,
         username: str,
         jwt_token: str,
-    ) -> int | None:
+    ) -> str | None:
         pass
 
     @abstractmethod
@@ -35,7 +35,7 @@ async def attach_command(
         job_id: str,
         username: str,
         jwt_token: str,
-    ) -> int | None:
+    ) -> str | None:
         pass
 
     @abstractmethod

src/lib/scheduler_clients/slurm/slurm_cli_client.py

@@ -66,7 +66,7 @@ async def submit_job(
         job_description: SlurmJobDescription,
         username: str,
         jwt_token: str,
-    ) -> int | None:
+    ) -> str | None:
         sbatch = SbatchCommand(job_description=job_description)
         return await self.__executed_ssh_cmd(
             username, jwt_token, sbatch, job_description.script
@@ -78,7 +78,7 @@ async def attach_command(
         job_id: str,
         username: str,
         jwt_token: str,
-    ) -> int | None:
+    ) -> None:
         srun = SrunCommand(command=command, job_id=job_id, overlap=True)
         return await self.__executed_ssh_cmd(username, jwt_token, srun)
 

src/lib/scheduler_clients/slurm/slurm_client.py

@@ -55,7 +55,7 @@ async def submit_job(
         job_description: SlurmJobDescription,
         username: str,
         jwt_token: str,
-    ) -> int | None:
+    ) -> str | None:
 
         if job_description.script_path:
             if self.slurm_cli_client:
@@ -74,7 +74,7 @@ async def submit_job(
 
     async def attach_command(
         self, command: str, job_id: str, username: str, jwt_token: str
-    ) -> int | None:
+    ) -> None:
         return await self.slurm_default_client.attach_command(
             command, job_id, username, jwt_token
         )
@@ -97,12 +97,14 @@ async def get_job_metadata(
         self, job_id: str, username: str, jwt_token: str
     ) -> List[SlurmJobMetadata]:
         if self.slurm_cli_client:
-            return await self.slurm_cli_client.get_job_metadata(job_id, username, jwt_token)
+            return await self.slurm_cli_client.get_job_metadata(
+                job_id, username, jwt_token
+            )
         else:
             raise HTTPException(
-                    status_code=501,
-                    detail="Scheduler can't handle this request when configured to use rest connection mode",
-                )
+                status_code=501,
+                detail="Scheduler can't handle this request when configured to use rest connection mode",
+            )
 
     async def get_nodes(self, username: str, jwt_token: str) -> List[SlurmNode] | None:
         return await self.slurm_default_client.get_nodes(username, jwt_token)

src/lib/scheduler_clients/slurm/slurm_rest_client.py

@@ -92,7 +92,7 @@ async def submit_job(
         job_description: SlurmJobDescription,
         username: str,
         jwt_token: str,
-    ) -> int | None:
+    ) -> str | None:
 
         client = await self.get_aiohttp_client()
         timeout = aiohttp.ClientTimeout(total=self.timeout)
@@ -132,15 +132,18 @@ async def submit_job(
             if response.status != status.HTTP_200_OK:
                 await _slurm_unexpected_response(response)
             job_submit_result = await response.json()
-        return job_submit_result["job_id"]
+        job_id = job_submit_result.get("job_id")
+        if job_id is None:
+            return None
+        return str(job_id)
 
     async def attach_command(
         self,
         command: str,
         job_id: str,
         username: str,
         jwt_token: str,
-    ) -> int | None:
+    ) -> None:
         pass
 
     async def get_job(

tests/mocked_api_responses/slurm_submit_job.json

@@ -14,7 +14,7 @@
         }
     },
     "errors": [],
-    "job_id": 42,
+    "job_id": "42",
     "step_id": "BATCH",
     "job_submit_user_msg": ""
 }