Merge pull request #1750 from buildtesters/lsf_queue_check_in_executors

Refactor code for LSF queue validation for executor check

buildtest/config.py

@@ -263,45 +263,16 @@ def _validate_lsf_executors(self):
         if not lsf.active():
             return
 
-        queue_list = []
-        valid_queue_state = "Open:Active"
-
-        record = lsf.queues()["RECORDS"]
-        # retrieve all queues from json record
-        for name in record:
-            queue_list.append(name["QUEUE_NAME"])
-
         # check all executors have defined valid queues and check queue state.
         for executor in lsf_executors:
             executor_name = f"{self.name()}.{executor_type}.{executor}"
             if self.is_executor_disabled(lsf_executors[executor]):
                 self.disabled_executors.append(executor_name)
                 continue
 
-            queue = lsf_executors[executor].get("queue")
-            # if queue field is defined check if its valid queue
-            if queue:
-                if queue not in queue_list:
-                    self.invalid_executors.append(executor_name)
-                    logger.error(
-                        f"'{queue}' is invalid LSF queue. Please select one of the following queues: {queue_list}"
-                    )
-                    continue
-
-                # check queue record for Status
-                for name in record:
-                    # skip record until we find matching queue
-                    if name["QUEUE_NAME"] != queue:
-                        continue
-
-                    queue_state = name["STATUS"]
-                    # if state not Open:Active we raise error
-                    if not queue_state == valid_queue_state:
-                        self.invalid_executors.append(executor_name)
-                        logger.error(
-                            f"'{queue}' is in state: {queue_state}. It must be in {valid_queue_state} state in order to accept jobs"
-                        )
-                        break
+            if not lsf.validate_queue(executor=lsf_executors[executor]):
+                self.invalid_executors.append(executor_name)
+                continue
 
             self.valid_executors[executor_type][executor_name] = {
                 "setting": lsf_executors[executor]

buildtest/scheduler/detection.py

@@ -261,6 +261,42 @@ def get_queues(self):
 
         return queues
 
+    def validate_queue(self, executor):
+        """This method will validate a LSF queue. We check if queue is available and in 'Open:Active' state.
+        The input is a dictionary containing the LSF executor configuration. If queue is not
+        found we return False.
+
+        Args:
+            executor (dict): The dictionary containing the LSF executor configuration.
+
+        Returns:
+            bool: True if queue is found and in 'Open:Active' state, False otherwise.
+        """
+
+        queue_name = executor["queue"]
+        queue_active_state = "Open:Active"
+
+        queue_list = [name["QUEUE_NAME"] for name in self._queues["RECORDS"]]
+        if queue_name not in queue_list:
+            return False
+
+        for record in self._queues["RECORDS"]:
+            # check queue record for Status
+
+            # skip record until we find matching queue
+            if record["QUEUE_NAME"] != queue_name:
+                continue
+
+            queue_state = record["STATUS"]
+            # if state not Open:Active we raise error
+            if not queue_state == queue_active_state:
+                self.logger.error(
+                    f"'{queue_name}' is in state: {queue_state}. It must be in {queue_active_state} state in order to accept jobs"
+                )
+                return False
+
+        return True
+
 
 class Cobalt(Scheduler):
     """The Cobalt class checks for Cobalt binaries and gets a list of Cobalt queues"""

docs/batch_support.rst

@@ -1146,7 +1146,7 @@ Let's run this example and notice that this job ran to completion but it was rep
 Torque
 -------
 
-Buildtest has support for running jobs on `Torque <https://adaptivecomputing.com/products/torque-resource-manager/>`_ scheduler. You must
+Buildtest has support for running jobs on `Torque <https://adaptivecomputing.com/cherry-services/torque-resource-manager/>`_ scheduler. You must
 define a :ref:`torque_executors` in your configuration file to use Torque scheduler. The ``#PBS`` directives can be specified using
 ``pbs`` property which is a list of PBS options that get inserted at top of script. Shown below is an example sleep job that will run on
 a single node for 5 seconds.

docs/configuring_buildtest/site_examples.rst

@@ -17,20 +17,24 @@ configuring compilers that is available on Perlmutter.
 Oak Ridge National Laboratory
 -----------------------------
 
-`Summit <https://docs.olcf.ornl.gov/systems/summit_user_guide.html>`_ is a training
-system for Summit at OLCF, which is using a IBM Load Sharing
-Facility (LSF) as their batch scheduler. Ascent has two
-queues **batch** and **test**. To declare LSF executors we define them under ``lsf``
+`Summit <https://docs.olcf.ornl.gov/systems/summit_user_guide.html>`_ is a IBM based system
+hosted at Oak Ridge Leadership Computing Facility (OLCF). The system uses IBM Load Sharing
+Facility (LSF) as their batch scheduler.
+
+The ``system`` keyword is used to define the name of system which in this example is named ``summit``. The
+``hostnames`` is used to specify a list of hostnames where buildtest can run in order to use this system configuration.
+
+The system comes with several queues, for the purposes of this example we define 3 executors
+that map to queues **batch** , **test** and **storage**. To declare LSF executors we define them under ``lsf``
 section within the ``executors`` section.
 
 The default batch configuration is defined in ``defaults``, for instance we set the fields ``pollinterval``, ``maxpendtime``
 and to **30s** and **300s** each. The field ``account`` is used to specify project account where all jobs will be charged. This can be
 customized to each site but and can be changed in the configuration file or overridden via command line ``buildtest build --account <ACCOUNT>``.
 
-
 .. literalinclude:: ../../tests/settings/summit.yml
    :language: yaml
-   :emphasize-lines: 19-23,37-39
+   :emphasize-lines: 2-5,19-23,37-43
 
 Argonne National Laboratory
 ---------------------------

tests/settings/summit.yml

@@ -37,6 +37,10 @@ system:
       lsf:
         batch:
           queue: batch
+        storage:
+          queue: storage
+        debug:
+          queue: debug
     compilers:
       find:
         gcc: ^(gcc)