feat: Support parallelly running for insights-engine (#3436)

Signed-off-by: Xiangce Liu <[email protected]>
(cherry picked from commit bbabfae)

insights/core/evaluators.py

@@ -53,9 +53,13 @@ def preprocess(self):
     def run_serial(self, graph=None):
         dr.run(graph or dr.COMPONENTS[dr.GROUPS.single], broker=self.broker)
 
-    def run_incremental(self, graph=None):
-        for _ in dr.run_incremental(graph or dr.COMPONENTS[dr.GROUPS.single], broker=self.broker):
-            pass
+    def run_incremental(self, graph=None, parallel=False):
+        components = graph or dr.COMPONENTS[dr.GROUPS.single]
+        if parallel:
+            with insights.get_pool(parallel, "insights-engine-pool", {"max_workers": None}) as pool:
+                dr.run_all(components, self.broker, pool)
+        else:
+            dr.run_all(components, self.broker)
 
     def format_response(self, response):
         """
@@ -70,10 +74,10 @@ def format_result(self, result):
         """
         return result
 
-    def process(self, graph=None):
+    def process(self, graph=None, parallel=False):
         with self:
             if self.incremental:
-                self.run_incremental(graph)
+                self.run_incremental(graph, parallel)
             else:
                 self.run_serial(graph)
         return self.get_response()