stepup-queue 1.0.2__tar.gz → 1.0.4__tar.gz

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: stepup-queue
-Version: 1.0.2
+Version: 1.0.4
 Summary: StepUp Queue integrates queued jobs into a StepUp workflow.
 Author-email: Toon Verstraelen <toon.verstraelen@ugent.be>
 License-Expression: GPL-3.0-or-later

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/docs/changelog.md

@@ -12,6 +12,22 @@ and this project adheres to [Effort-based Versioning](https://jacobtomlinson.dev
 
 (no changes yet)
 
+## [1.0.4][] - 2025-05-21 {: #v1.0.4 }
+
+### Fixed
+
+- Minor typo fix in slurm wrapper script.
+- Improved example perpetual workflow job script.
+
+## [1.0.3][] - 2025-05-16 {: #v1.0.3 }
+
+### Fixed
+
+- Fixed errors in the example job scripts.
+- Improved handling of `scontrol` failures.
+
+### Added
+
 ## [1.0.2][] - 2025-05-14 {: #v1.0.2 }
 
 ### Added
@@ -38,6 +54,8 @@ It was adapted to integrate well with StepUp Core 3.
 This release also features the `stepup canceljobs` tool, which was not present in Parman.
 
 [Unreleased]: https://github.com/reproducible-reporting/stepup-queue
+[1.0.4]: https://github.com/reproducible-reporting/stepup-queue/releases/tag/v1.0.4
+[1.0.3]: https://github.com/reproducible-reporting/stepup-queue/releases/tag/v1.0.3
 [1.0.2]: https://github.com/reproducible-reporting/stepup-queue/releases/tag/v1.0.2
 [1.0.1]: https://github.com/reproducible-reporting/stepup-queue/releases/tag/v1.0.1
 [1.0.0]: https://github.com/reproducible-reporting/stepup-queue/releases/tag/v1.0.0

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/docs/examples/slurm-basic/dynamic-template.sh

@@ -1,7 +1,7 @@
 #!/usr/bin/env bash
 #SBATCH --job-name 'dyn{{ field }}'
 #SBATCH --nodes=1
-#SBATCH --num-tasks=1
+#SBATCH --ntasks=1
 #SBATCH --cpus-per-task=1
 
 echo "Hello from dynamic job {{ field }}"

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/docs/examples/slurm-basic/fail/slurmjob.sh

@@ -1,7 +1,7 @@
 #!/usr/bin/env bash
 #SBATCH --job-name fail
 #SBATCH --nodes=1
-#SBATCH --num-tasks=1
+#SBATCH --ntasks=1
 #SBATCH --cpus-per-task=1
 
 echo "This job will fail"

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/docs/examples/slurm-basic/pass/slurmjob.py

@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 #SBATCH --job-name pass
 #SBATCH --nodes=1
-#SBATCH --num-tasks=1
+#SBATCH --ntasks=1
 #SBATCH --cpus-per-task=1
 
 from time import sleep

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/docs/examples/slurm-perpetual/step1/slurmjob.sh

@@ -1,7 +1,7 @@
 #!/usr/bin/env bash
 #SBATCH --job-name step1
 #SBATCH --nodes=1
-#SBATCH --num-tasks=1
+#SBATCH --ntasks=1
 #SBATCH --cpus-per-task=1
 #SBATCH --time=00:02:00
 

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/docs/examples/slurm-perpetual/step2/slurmjob.sh

@@ -1,7 +1,7 @@
 #!/usr/bin/env bash
 #SBATCH --job-name step2
 #SBATCH --nodes=1
-#SBATCH --num-tasks=1
+#SBATCH --ntasks=1
 #SBATCH --cpus-per-task=1
 
 #SBATCH --time=00:02:00

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/docs/examples/slurm-perpetual/workflow.sh

@@ -1,12 +1,13 @@
 #!/usr/bin/env bash
-#SBATCH --job-name perpetual-workflow
+#SBATCH --job-name stepup
 #SBATCH --nodes=1
-#SBATCH --num-tasks=1
+#SBATCH --ntasks=1
 #SBATCH --cpus-per-task=1
-#SBATCH --output=workflow-%j.out
+#SBATCH --output=stepup-%j.out
 #SBATCH --time=00:01:00
 
-# In production, --time=00:12:00 is a reasonable time limit.
+# In production, --time=12:00:00 is a reasonable time limit.
+echo "StepUp workflow job starts:" $(date)
 
 # If needed, load required modules and activate a relevant virtual environment.
 # For example:
@@ -24,10 +25,10 @@ trap 'rm -rv "$STEPUP_QUEUE_FLAG_DIR"' EXIT
 # The second will forcefully terminate remaining running steps.
 echo "Starting background process to monitor wall time."
 (
-    sleep 30;  # In production, 39600 seconds is reasonable.
-    touch ${STEPUP_QUEUE_FLAG_DIR}/resubmit;
-    stepup shutdown;
-    sleep 10;  # In production, 300 seconds is reasonable.
+    sleep 30  # In production, wall time minus 1800 seconds (half hour) is reasonable.
+    touch ${STEPUP_QUEUE_FLAG_DIR}/resubmit
+    stepup shutdown
+    sleep 10  # In production, 300 seconds (5 minutes) is reasonable.
     stepup shutdown
 ) &
 BGPID=$!
@@ -48,5 +49,7 @@ if [ -f ${STEPUP_QUEUE_FLAG_DIR}/resubmit ]; then
     echo "Resubmitting job script to let StepUp finalize the workflow."
     sbatch workflow.sh
 else
-    echo "Stepup was stopped gracefully."
+    echo "Stepup stopped by itself."
 fi
+
+echo "StepUp workflow job ends:" $(date)

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/stepup/queue/sbatch.py

@@ -177,10 +177,10 @@ def _read_or_poll_status(
         # Call scontrol and parse its response.
         rndsleep()
         status_time, status = get_status(work_thread, jobid, cluster)
-        if status is not None and status != last_status:
+        if status != last_status:
             log_step(path_log, status)
     done = (status_time > submit_time + TIME_MARGIN) and (
-        status not in ["PENDING", "CONFIGURING", "RUNNING"]
+        status not in ["PENDING", "CONFIGURING", "RUNNING", "invalid"]
     )
     return status, done
 
@@ -235,7 +235,7 @@ chmod +x '{job_script}'
 ./'{job_script}'
 RETURN_CODE=$?
 echo $RETURN_CODE > slurmjob.ret
-exot $RETURN_CODE
+exit $RETURN_CODE
 """
 
 
@@ -281,7 +281,7 @@ def parse_sbatch(stdout: str) -> tuple[int, str | None]:
     raise ValueError(f"Cannot parse sbatch output: {stdout}")
 
 
-def get_status(work_thread: WorkThread, jobid: int, cluster: str | None) -> str | None:
+def get_status(work_thread: WorkThread, jobid: int, cluster: str | None) -> str:
     """Load cached scontrol output or run scontrol if outdated.
 
     Parameters
@@ -296,10 +296,9 @@ def get_status(work_thread: WorkThread, jobid: int, cluster: str | None) -> str
     Returns
     -------
     status
-        A status reported by scontrol.
-        The "Invalid" status is returned when scontrol fails to find the jobid.
-        None is returned when scontrol fails in a way that is safe to ignore.
-        (Try again later.)
+        A status reported by scontrol,
+        or `invalid` if scontrol failed (retry scontrol later),
+        or `unlisted` if the job is not found (probably ended long ago).
     """
     # Load cached output or run again
     command = "scontrol show job"
@@ -309,11 +308,17 @@ def get_status(work_thread: WorkThread, jobid: int, cluster: str | None) -> str
     else:
         command += f" --cluster={cluster}"
         path_out /= f"sbatch_wait.{cluster}.out"
-    status_time, scontrol_out = cached_run(work_thread, command, path_out, CACHE_TIMEOUT)
+    status_time, scontrol_out, returncode = cached_run(
+        work_thread, command, path_out, CACHE_TIMEOUT
+    )
+    if returncode != 0:
+        return status_time, "invalid"
     return status_time, parse_scontrol_out(scontrol_out, jobid)
 
 
-def cached_run(work_thread: WorkThread, command: str, path_out: Path, cache_timeout) -> str:
+def cached_run(
+    work_thread: WorkThread, command: str, path_out: Path, cache_timeout
+) -> tuple[float, str, int]:
     """Execute a command if its previous output is outdated.
 
     Parameters
@@ -329,8 +334,12 @@ def cached_run(work_thread: WorkThread, command: str, path_out: Path, cache_time
 
     Returns
     -------
+    cache_time
+        The time when the command was last executed.
     stdout
         The output of the file, either new or cached.
+    returncode
+        The return code of the (cached) command.
 
     Notes
     -----
@@ -345,7 +354,7 @@ def cached_run(work_thread: WorkThread, command: str, path_out: Path, cache_time
         fcntl.lockf(fh, fcntl.LOCK_EX)
         fh.seek(0)
         header = fh.read(CACHE_HEADER_LENGTH)
-        cache_time, _ = parse_cache_header(header)
+        cache_time, returncode = parse_cache_header(header)
         if cache_time is None or time.time() > cache_time + cache_timeout:
             returncode, stdout, _ = work_thread.runsh(command)
             # Go the the beginning of the file before truncating.
@@ -358,8 +367,8 @@ def cached_run(work_thread: WorkThread, command: str, path_out: Path, cache_time
             fh.write(stdout)
             fh.flush()
             os.fsync(fh.fileno())
-            return cache_time, stdout
-        return cache_time, fh.read()
+            return cache_time, stdout, returncode
+        return cache_time, fh.read(), returncode
 
 
 def make_cache_header(cache_time: float, returncode: int):
@@ -386,7 +395,7 @@ def parse_cache_header(header: str) -> tuple[float, int]:
 CACHE_HEADER_LENGTH = len(make_cache_header(time.time(), 0))
 
 
-def parse_scontrol_out(scontrol_out: str, jobid: int) -> str | None:
+def parse_scontrol_out(scontrol_out: str, jobid: int) -> str:
     """Get the job state for a specific from from the output of ``scontrol show job``.
 
     Parameters
@@ -399,10 +408,12 @@ def parse_scontrol_out(scontrol_out: str, jobid: int) -> str | None:
     Returns
     -------
     jobstate
-        The status of the job, or None of the job cannot be found.
+        The status of the job. This can be:
+
+        - Any of the SLURM job states.
+        - `unlisted` if the job cannot be found,
+          which practically means it has ended long ago.
     """
-    if scontrol_out == SCONTROL_FAILED:
-        return "Invalid"
     match = re.search(
         f"JobId={jobid}.*?JobState=(?P<state>[A-Z]+)",
         scontrol_out,
@@ -410,4 +421,4 @@ def parse_scontrol_out(scontrol_out: str, jobid: int) -> str | None:
     )
     if match is not None:
         return match.group("state")
-    return None
+    return "unlisted"

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/stepup_queue.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: stepup-queue
-Version: 1.0.2
+Version: 1.0.4
 Summary: StepUp Queue integrates queued jobs into a StepUp workflow.
 Author-email: Toon Verstraelen <toon.verstraelen@ugent.be>
 License-Expression: GPL-3.0-or-later

{stepup_queue-1.0.2 → stepup_queue-1.0.4}/tests/test_sbatch.py

@@ -56,15 +56,17 @@ def test_parse_sbatch():
 def test_cached_run(path_tmp: Path):
     path_out = path_tmp / "date.txt"
     work_thread = WorkThread("<test>")
-    cache_time1, out1 = cached_run(work_thread, "date", path_out, 1)
-    cache_time2, out2 = cached_run(work_thread, "date", path_out, 10)
+    cache_time1, out1, ret1 = cached_run(work_thread, "date", path_out, 1)
+    cache_time2, out2, ret2 = cached_run(work_thread, "date", path_out, 10)
     assert cache_time1 == pytest.approx(cache_time2, 1e-4)
     assert out1 != ""
     assert out1 == out2
+    assert ret1 == ret2
     time.sleep(2)
-    cache_time3, out3 = cached_run(work_thread, "date", path_out, 1)
+    cache_time3, out3, ret3 = cached_run(work_thread, "date", path_out, 1)
     assert abs(cache_time1 - cache_time3) > 0.5
     assert out1 != out3
+    assert ret1 == ret3
 
 
 SCONTROL_OUT = """\