Implemented old results pruning

openscap_daemon/config.py

@@ -44,6 +44,8 @@ def __init__(self):
         self.cve_feeds_dir = \
             os.path.join("/", "var", "lib", "oscapd", "cve_feeds")
         self.jobs = 4
+        # -2 means never prune old results
+        self.max_results_to_keep = -2
 
         # Tools section
         self.oscap_path = ""
@@ -224,6 +226,11 @@ def absolutize(path):
         except (configparser.NoOptionError, configparser.NoSectionError):
             pass
 
+        try:
+            self.max_results_to_keep = config.getint("General", "max-results-to-keep")
+        except (configparser.NoOptionError, configparser.NoSectionError):
+            pass
+
         # Tools section
         try:
             self.oscap_path = absolutize(config.get("Tools", "oscap"))
@@ -295,6 +302,7 @@ def save_as(self, config_file):
         config.set("General", "work-in-progress-dir", str(self.work_in_progress_dir))
         config.set("General", "cve-feeds-dir", str(self.cve_feeds_dir))
         config.set("General", "jobs", str(self.jobs))
+        config.set("General", "max-results-to-keep", str(self.max_results_to_keep))
 
         config.add_section("Tools")
         config.set("Tools", "oscap", str(self.oscap_path))

openscap_daemon/task.py

@@ -169,6 +169,11 @@ def __init__(self):
         self.title = None
         self.evaluation_spec = evaluation_spec.EvaluationSpec()
 
+        # How many results should we keep before pruning old results
+        # -1 means use the default from config
+        # -2 means never prune any results
+        self.max_results_to_keep = -1
+
         self.schedule = Schedule()
         # If True, this task will be evaluated once without affecting the
         # schedule. This feature is important for test runs. This variable does
@@ -183,6 +188,9 @@ def __str__(self):
         ret += "- ID: \t%i\n" % (self.id_)
         ret += "- title: \t%s\n" % (self.title)
         ret += str(self.evaluation_spec) + "\n"
+        ret += "- max results to keep: \t%s\n" % \
+            ("default" if self.max_results_to_keep == -1
+             else str(self.max_results_to_keep))
         ret += "- schedule:\n"
         ret += "  - not before: \t%s\n" % (self.schedule.not_before)
         ret += "  - repeat after: \t%s\n" % (self.schedule.repeat_after)
@@ -205,6 +213,7 @@ def is_equivalent_to(self, other):
         return \
             self.evaluation_spec.is_equivalent_to(other.evaluation_spec) and \
             self.title == other.title and \
+            self.max_results_to_keep == other.max_results_to_keep and \
             self.schedule.is_equivalent_to(other.schedule) and \
             self.run_outside_schedule_once == other.run_outside_schedule_once
 
@@ -229,6 +238,9 @@ def load_from_xml_element(self, root, config_file):
             et_helpers.get_element(root, "evaluation_spec")
         )
 
+        self.max_results_to_keep = \
+            int(et_helpers.get_element_text(root, "max-results-to-keep", "-1"))
+
         self.schedule = Schedule()
         self.schedule.load_from_xml_element(
             et_helpers.get_element(root, "schedule")
@@ -259,6 +271,11 @@ def to_xml_element(self):
         evaluation_spec_element = self.evaluation_spec.to_xml_element()
         root.append(evaluation_spec_element)
 
+        if self.max_results_to_keep != -1:
+            max_results_element = ElementTree.Element("max-results-to-keep")
+            max_results_element.text = str(self.max_results_to_keep)
+            root.append(max_results_element)
+
         schedule_element = self.schedule.to_xml_element()
         root.append(schedule_element)
 
@@ -356,12 +373,14 @@ def remove_result(self, result_id, config):
         )
 
         logging.debug(
-            "Removing ARF of result '%i' of task '%i', expected path '%s'.",
-            result_id, self.id_, result_path
+            "Removing ARF of result '%s' of task '%i', expected path '%s'.",
+            str(result_id), self.id_, result_path
         )
 
         shutil.rmtree(result_path, False)
-        logging.info("Removed result '%i' of task '%i'.", result_id, self.id_)
+        logging.info(
+            "Removed result '%s' of task '%i'.", str(result_id), self.id_
+        )
 
     def get_next_update_time(self, reference_datetime, log=False):
         if not self.enabled:
@@ -404,6 +423,24 @@ def should_be_updated(self, reference_datetime, log=False):
 
         return False
 
+    def prune_old_results(self, config):
+        max_results_to_keep = self.max_results_to_keep
+        if max_results_to_keep == -1:
+            max_results_to_keep = config.max_results_to_keep
+
+        if max_results_to_keep < 0:
+            # pruning is disabled
+            return
+
+        result_ids = self.list_result_ids(config.results_dir)
+        result_ids_to_remove = result_ids[max_results_to_keep:]
+
+        if result_ids_to_remove:
+            logging.info("Pruning old results of task '%i'...", self.id_)
+
+            for result_id in reversed(result_ids_to_remove):
+                self.remove_result(result_id, config)
+
     def update(self, reference_datetime, config):
         """Figures out if the task should be run right now, alters the schedule
         values accordingly.
@@ -447,6 +484,9 @@ def update(self, reference_datetime, config):
                 else:
                     self.run_outside_schedule_once = False
 
+                # we have one extra result, let's prune old results
+                self.prune_old_results(config)
+
     def generate_guide(self, config):
         return self.evaluation_spec.generate_guide(config)
 

tests/data_dir_template/config.ini

@@ -4,6 +4,7 @@ results-dir=./results
 work-in-progress-dir=./work_in_progress
 cve-feeds-dir=./cve_feeds
 jobs=4
+max-results-to-keep=100
 
 [Tools]
 oscap=