From 400d90da07ecabeffca17bf9ca7bbe85339345cf Mon Sep 17 00:00:00 2001
From: Dmitry Mescheryakov <dmescheryakov@mirantis.com>
Date: Tue, 15 Apr 2014 18:22:50 +0400
Subject: [PATCH] Avoid deleting transient cluster before job is started

In order to run job on a transient cluster client needs to execute
two commands:
 * create cluster with is_transient=true
 * run job on it in a regular manner

We terminate unneeded transient clusters in a periodic job, which
terminates cluster if cluster is transient and no job is running on it
at a time. We also do not terminate cluster if its lifetime is smaller
then config parameter min_transient_cluster_active_time. For some
reason the parameter is set to 0 by default, which could cause
premature cluster termination if periodic task runs between
cluster creation and job execution.

Also added unit tests to verify min_transient_cluster_active_time.

Change-Id: I5330a969ea7cd81ec2759a7fe32bab4a5de3fb4c
---
 etc/sahara/sahara.conf.sample              |  2 +-
 sahara/service/periodic.py                 |  2 +-
 sahara/tests/unit/service/test_periodic.py | 63 +++++++++++++++++-----
 3 files changed, 52 insertions(+), 15 deletions(-)

diff --git a/etc/sahara/sahara.conf.sample b/etc/sahara/sahara.conf.sample
index 7d14ed5b..a14c6b2e 100644
--- a/etc/sahara/sahara.conf.sample
+++ b/etc/sahara/sahara.conf.sample
@@ -258,7 +258,7 @@
 # Minimal "lifetime" in seconds for a transient cluster.
 # Cluster is guarantied to be "alive" within this time period.
 # (integer value)
-#min_transient_cluster_active_time=0
+#min_transient_cluster_active_time=30
 
 
 #
diff --git a/sahara/service/periodic.py b/sahara/service/periodic.py
index ab897853..a60d2449 100644
--- a/sahara/service/periodic.py
+++ b/sahara/service/periodic.py
@@ -44,7 +44,7 @@ periodic_opts = [
                help='Max interval size between periodic tasks execution in '
                     'seconds'),
     cfg.IntOpt('min_transient_cluster_active_time',
-               default=0,
+               default=30,
                help='Minimal "lifetime" in seconds for a transient cluster. '
                     'Cluster is guarantied to be "alive" within this time '
                     'period.'),
diff --git a/sahara/tests/unit/service/test_periodic.py b/sahara/tests/unit/service/test_periodic.py
index 78d29eb1..23c640a7 100644
--- a/sahara/tests/unit/service/test_periodic.py
+++ b/sahara/tests/unit/service/test_periodic.py
@@ -51,23 +51,20 @@ class TestPeriodicBack(base.SaharaWithDbTestCase):
         get_job_status.assert_has_calls([mock.call(u'2'),
                                          mock.call(u'3')])
 
-    @mock.patch('sahara.service.edp.job_manager.get_job_status')
+    @mock.patch('sahara.openstack.common.timeutils.utcnow')
     @mock.patch('sahara.service.api.terminate_cluster')
-    def test_cluster_terminate(self, terminate_cluster, get_job_status):
-        self.override_config("use_identity_api_v3", True)
+    def test_cluster_terminate(self, terminate_cluster, utcnow):
+
+        utcnow.return_value = datetime.datetime(2005, 2, 1, 0, 0)
+
         ctx = context.ctx()
         job = self.api.job_create(ctx, te.SAMPLE_JOB)
         ds = self.api.data_source_create(ctx, te.SAMPLE_DATA_SOURCE)
-        c = tc.SAMPLE_CLUSTER.copy()
-        c["status"] = "Active"
-        c["id"] = "1"
-        c["name"] = "1"
-        c['updated_at'] = timeutils.utcnow()
-        self.api.cluster_create(ctx, c)
-        c["id"] = "2"
-        c["name"] = "2"
-        self.api.cluster_create(ctx, c)
-        self._create_job_execution({"end_time": datetime.datetime.now(),
+
+        self._make_cluster('1')
+        self._make_cluster('2')
+
+        self._create_job_execution({"end_time": timeutils.utcnow(),
                                     "id": 1,
                                     "cluster_id": "1"},
                                    job, ds, ds)
@@ -79,10 +76,50 @@ class TestPeriodicBack(base.SaharaWithDbTestCase):
                                     "id": 3,
                                     "cluster_id": "2"},
                                    job, ds, ds)
+
+        utcnow.return_value = datetime.datetime(2005, 2, 1, 0, 1)
+
         p.SaharaPeriodicTasks().terminate_unneeded_clusters(None)
         self.assertEqual(terminate_cluster.call_count, 1)
         terminate_cluster.assert_has_calls([mock.call(u'1')])
 
+    @mock.patch('sahara.openstack.common.timeutils.utcnow')
+    @mock.patch('sahara.service.api.terminate_cluster')
+    def test_cluster_not_killed_too_early(self, terminate_cluster, utcnow):
+
+        utcnow.return_value = datetime.datetime(2005, 2, 1, second=0)
+
+        self._make_cluster('1')
+
+        utcnow.return_value = datetime.datetime(2005, 2, 1, second=20)
+
+        p.SaharaPeriodicTasks().terminate_unneeded_clusters(None)
+        self.assertEqual(terminate_cluster.call_count, 0)
+
+    @mock.patch('sahara.openstack.common.timeutils.utcnow')
+    @mock.patch('sahara.service.api.terminate_cluster')
+    def test_cluster_killed_in_time(self, terminate_cluster, utcnow):
+
+        utcnow.return_value = datetime.datetime(2005, 2, 1, second=0)
+
+        self._make_cluster('1')
+
+        utcnow.return_value = datetime.datetime(2005, 2, 1, second=40)
+
+        p.SaharaPeriodicTasks().terminate_unneeded_clusters(None)
+        self.assertEqual(terminate_cluster.call_count, 1)
+        terminate_cluster.assert_has_calls([mock.call(u'1')])
+
+    def _make_cluster(self, id_name):
+        ctx = context.ctx()
+
+        c = tc.SAMPLE_CLUSTER.copy()
+        c["status"] = "Active"
+        c["id"] = id_name
+        c["name"] = id_name
+        c['updated_at'] = timeutils.utcnow()
+        self.api.cluster_create(ctx, c)
+
     def _create_job_execution(self, values, job, input, output):
         values.update({"job_id": job['id'],
                        "input_id": input['id'],