Fatal if unable to load script or env when building a launch job msg.

If a file that slurmctld wrote disappears (job script or job environment),
then the StateSaveLocation is in unexpected state and Slurm should fatal out
immediately before corrupting anything else. Give user the option to
explicitly ignore the error at startup at the loss of the job.

Bug 7783.

Author: naterini

NEWS

@@ -38,6 +38,8 @@ documents those changes that are of interest to users and administrators.
     conflicts on development systems running multiple Slurm instances.
  -- Return a newly added ESLURM_DEFER error and set a job state reason to
     FAIL_DEFER for immediate alloc requests if defer in SchedulerParameters.
+ -- Make slurmctld fatal if unable to load a script or a job environment when
+    building the launch job message.
 
 * Changes in Slurm 19.05.5
 ==========================

src/slurmctld/job_scheduler.c

@@ -2040,6 +2040,7 @@ static void _split_env(batch_job_launch_msg_t *launch_msg_ptr)
 static batch_job_launch_msg_t *_build_launch_job_msg(job_record_t *job_ptr,
 						     uint16_t protocol_version)
 {
+	char *fail_why = NULL;
 	batch_job_launch_msg_t *launch_msg_ptr;
 
 	/* Initialization of data structures */
@@ -2054,18 +2055,10 @@ static batch_job_launch_msg_t *_build_launch_job_msg(job_record_t *job_ptr,
 	launch_msg_ptr->gid = job_ptr->group_id;
 
 	if (!(launch_msg_ptr->script_buf = get_job_script(job_ptr))) {
-		error("Can not find batch script, aborting batch %pJ",
-		      job_ptr);
-		/* FIXME: This is a kludge, but this event indicates a missing
-		 * batch script and should never happen. We are too deep into
-		 * the job launch to gracefully clean up here. */
-		slurm_free_job_launch_msg(launch_msg_ptr);
-		job_complete(job_ptr->job_id, slurmctld_conf.slurm_user_id,
-			     true, false, 0);
-		return NULL;
+		fail_why = "Unable to load job batch script";
+		goto job_failed;
 	}
 
-
 	launch_msg_ptr->ntasks = job_ptr->details->num_tasks;
 	launch_msg_ptr->alias_list = xstrdup(job_ptr->alias_list);
 	launch_msg_ptr->nodes = xstrdup(job_ptr->nodes);
@@ -2111,13 +2104,10 @@ static batch_job_launch_msg_t *_build_launch_job_msg(job_record_t *job_ptr,
 	launch_msg_ptr->environment = get_job_env(job_ptr,
 						  &launch_msg_ptr->envc);
 	if (launch_msg_ptr->environment == NULL) {
-		error("%s: environment missing or corrupted aborting %pJ",
-		      __func__, job_ptr);
-		slurm_free_job_launch_msg(launch_msg_ptr);
-		job_complete(job_ptr->job_id, slurmctld_conf.slurm_user_id,
-			     false, true, 0);
-		return NULL;
+		fail_why = "Unable to load job environment";
+		goto job_failed;
 	}
+
 	_split_env(launch_msg_ptr);
 	launch_msg_ptr->job_mem = job_ptr->details->pn_min_memory;
 	launch_msg_ptr->num_cpu_groups = job_ptr->job_resrcs->cpu_array_cnt;
@@ -2146,7 +2136,26 @@ static batch_job_launch_msg_t *_build_launch_job_msg(job_record_t *job_ptr,
 	launch_msg_ptr->account = xstrdup(job_ptr->account);
 	launch_msg_ptr->resv_name = xstrdup(job_ptr->resv_name);
 
+	xassert(!fail_why);
 	return launch_msg_ptr;
+
+job_failed:
+	/* fatal or kill the job as it can never be recovered */
+	if (!ignore_state_errors)
+		fatal("%s: %s for %pJ. Check file system serving StateSaveLocation as that directory may be missing or corrupted. Start with '-i' to ignore this error and kill the afflicted jobs.",
+		      __func__, fail_why, job_ptr);
+
+	error("%s: %s for %pJ. %pJ will be killed due to system error.",
+	      __func__, fail_why, job_ptr, job_ptr);
+	xfree(job_ptr->state_desc);
+	job_ptr->state_desc = xstrdup(fail_why);
+	job_ptr->state_reason_prev = job_ptr->state_reason;
+	job_ptr->state_reason = FAIL_SYSTEM;
+	slurm_free_job_launch_msg(launch_msg_ptr);
+	/* ignore the return as job is in an unknown state anyway */
+	job_complete(job_ptr->job_id, slurmctld_conf.slurm_user_id, false,
+		     false, 1);
+	return NULL;
 }
 
 /* Validate the job is ready for launch