EC2 - Spot Requests improvements (#5087)

moto/ec2/_models/instances.py

@@ -99,8 +99,8 @@ class Instance(TaggedEC2Resource, BotoInstance, CloudFormationModel):
         self.launch_time = utc_date_and_time()
         self.ami_launch_index = kwargs.get("ami_launch_index", 0)
         self.disable_api_termination = kwargs.get("disable_api_termination", False)
-        self.instance_initiated_shutdown_behavior = kwargs.get(
+        self.instance_initiated_shutdown_behavior = (
-            "instance_initiated_shutdown_behavior", "stop"
+            kwargs.get("instance_initiated_shutdown_behavior") or "stop"
         )
         self.sriov_net_support = "simple"
         self._spot_fleet_id = kwargs.get("spot_fleet_id", None)

moto/ec2/_models/spot_requests.py

@@ -38,6 +38,7 @@ class SpotInstanceRequest(BotoSpotRequest, TaggedEC2Resource):
         subnet_id,
         tags,
         spot_fleet_id,
+        instance_interruption_behaviour,
         **kwargs,
     ):
         super().__init__(**kwargs)
@@ -46,12 +47,20 @@ class SpotInstanceRequest(BotoSpotRequest, TaggedEC2Resource):
         self.launch_specification = ls
         self.id = spot_request_id
         self.state = "open"
+        self.status = "pending-evaluation"
+        self.status_message = "Your Spot request has been submitted for review, and is pending evaluation."
+        if price:
+            price = float(price)
+            price = "{0:.6f}".format(price)  # round up/down to 6 decimals
         self.price = price
         self.type = spot_instance_type
         self.valid_from = valid_from
         self.valid_until = valid_until
         self.launch_group = launch_group
         self.availability_zone_group = availability_zone_group
+        self.instance_interruption_behaviour = (
+            instance_interruption_behaviour or "terminate"
+        )
         self.user_data = user_data  # NOT
         ls.kernel = kernel_id
         ls.ramdisk = ramdisk_id
@@ -62,7 +71,9 @@ class SpotInstanceRequest(BotoSpotRequest, TaggedEC2Resource):
         ls.monitored = monitoring_enabled
         ls.subnet_id = subnet_id
         self.spot_fleet_id = spot_fleet_id
-        self.tags = tags
+        tag_map = tags.get("spot-instances-request", {})
+        self.add_tags(tag_map)
+        self.all_tags = tags
 
         if security_groups:
             for group_name in security_groups:
@@ -75,6 +86,9 @@ class SpotInstanceRequest(BotoSpotRequest, TaggedEC2Resource):
             ls.groups.append(default_group)
 
         self.instance = self.launch_instance()
+        self.state = "active"
+        self.status = "fulfilled"
+        self.status_message = ""
 
     def get_filter_value(self, filter_name):
         if filter_name == "state":
@@ -95,7 +109,7 @@ class SpotInstanceRequest(BotoSpotRequest, TaggedEC2Resource):
             security_group_names=[],
             security_group_ids=self.launch_specification.groups,
             spot_fleet_id=self.spot_fleet_id,
-            tags=self.tags,
+            tags=self.all_tags,
             lifecycle="spot",
         )
         instance = reservation.instances[0]
@@ -128,6 +142,7 @@ class SpotRequestBackend(object, metaclass=Model):
         subnet_id,
         tags=None,
         spot_fleet_id=None,
+        instance_interruption_behaviour=None,
     ):
         requests = []
         tags = tags or {}
@@ -154,6 +169,7 @@ class SpotRequestBackend(object, metaclass=Model):
                 subnet_id,
                 tags,
                 spot_fleet_id,
+                instance_interruption_behaviour,
             )
             self.spot_instance_requests[spot_request_id] = request
             requests.append(request)
@@ -216,6 +232,7 @@ class SpotFleetRequest(TaggedEC2Resource, CloudFormationModel):
         allocation_strategy,
         launch_specs,
         launch_template_config,
+        instance_interruption_behaviour,
     ):
 
         self.ec2_backend = ec2_backend
@@ -224,6 +241,9 @@ class SpotFleetRequest(TaggedEC2Resource, CloudFormationModel):
         self.target_capacity = int(target_capacity)
         self.iam_fleet_role = iam_fleet_role
         self.allocation_strategy = allocation_strategy
+        self.instance_interruption_behaviour = (
+            instance_interruption_behaviour or "terminate"
+        )
         self.state = "active"
         self.fulfilled_capacity = 0.0
 
@@ -404,6 +424,7 @@ class SpotFleetBackend(object):
         allocation_strategy,
         launch_specs,
         launch_template_config=None,
+        instance_interruption_behaviour=None,
     ):
 
         spot_fleet_request_id = random_spot_fleet_request_id()
@@ -416,6 +437,7 @@ class SpotFleetBackend(object):
             allocation_strategy,
             launch_specs,
             launch_template_config,
+            instance_interruption_behaviour,
         )
         self.spot_fleet_requests[spot_fleet_request_id] = request
         return request

moto/ec2/responses/spot_fleets.py

@@ -47,6 +47,9 @@ class SpotFleets(BaseResponse):
         target_capacity = spot_config["TargetCapacity"]
         iam_fleet_role = spot_config["IamFleetRole"]
         allocation_strategy = spot_config["AllocationStrategy"]
+        instance_interruption_behaviour = spot_config.get(
+            "InstanceInterruptionBehavior"
+        )
 
         launch_specs = spot_config.get("LaunchSpecifications")
         launch_template_config = list(
@@ -63,6 +66,7 @@ class SpotFleets(BaseResponse):
             allocation_strategy=allocation_strategy,
             launch_specs=launch_specs,
             launch_template_config=launch_template_config,
+            instance_interruption_behaviour=instance_interruption_behaviour,
         )
 
         template = self.response_template(REQUEST_SPOT_FLEET_TEMPLATE)

moto/ec2/responses/spot_instances.py

@@ -62,6 +62,10 @@ class SpotInstances(EC2BaseResponse):
         ramdisk_id = self._get_param("LaunchSpecification.RamdiskId")
         monitoring_enabled = self._get_param("LaunchSpecification.Monitoring.Enabled")
         subnet_id = self._get_param("LaunchSpecification.SubnetId")
+        instance_interruption_behaviour = self._get_param(
+            "InstanceInterruptionBehavior"
+        )
+        tags = self._parse_tag_specification()
 
         if self.is_not_dryrun("RequestSpotInstance"):
             requests = self.ec2_backend.request_spot_instances(
@@ -82,6 +86,8 @@ class SpotInstances(EC2BaseResponse):
                 ramdisk_id=ramdisk_id,
                 monitoring_enabled=monitoring_enabled,
                 subnet_id=subnet_id,
+                instance_interruption_behaviour=instance_interruption_behaviour,
+                tags=tags,
             )
 
             template = self.response_template(REQUEST_SPOT_INSTANCES_TEMPLATE)
@@ -98,9 +104,9 @@ REQUEST_SPOT_INSTANCES_TEMPLATE = """<RequestSpotInstancesResponse xmlns="http:/
       <type>{{ request.type }}</type>
       <state>{{ request.state }}</state>
       <status>
-        <code>pending-evaluation</code>
+        <code>{{ request.status }}</code>
         <updateTime>2015-01-01T00:00:00.000Z</updateTime>
-        <message>Your Spot request has been submitted for review, and is pending evaluation.</message>
+        <message>{{ request.status_message }}</message>
       </status>
       <instanceId>{{ request.instance_id }}</instanceId>
       <availabilityZoneGroup>{{ request.availability_zone_group }}</availabilityZoneGroup>
@@ -153,11 +159,11 @@ DESCRIBE_SPOT_INSTANCES_TEMPLATE = """<DescribeSpotInstanceRequestsResponse xmln
       <type>{{ request.type }}</type>
       <state>{{ request.state }}</state>
       <status>
-        <code>pending-evaluation</code>
+        <code>{{ request.status }}</code>
         <updateTime>2015-01-01T00:00:00.000Z</updateTime>
-        <message>Your Spot request has been submitted for review, and is pending evaluation.</message>
+        <message>{{ request.status_message }}</message>
       </status>
-      <instanceId>{{ request.instance_id }}</instanceId>
+      <instanceId>{{ request.instance.id }}</instanceId>
       {% if request.availability_zone_group %}
         <availabilityZoneGroup>{{ request.availability_zone_group }}</availabilityZoneGroup>
       {% endif %}
@@ -217,6 +223,7 @@ DESCRIBE_SPOT_INSTANCES_TEMPLATE = """<DescribeSpotInstanceRequestsResponse xmln
         <validUntil>{{ request.valid_until }}</validUntil>
       {% endif %}
       <productDescription>Linux/UNIX</productDescription>
+      <instanceInterruptionBehavior>{{ request.instance_interruption_behaviour }}</instanceInterruptionBehavior>
     </item>
     {% endfor %}
   </spotInstanceRequestSet>

tests/terraformtests/etc/0002-EC2-reduce-wait-times.patch

@@ -342,7 +342,7 @@ index 49e4909b3a..731a37f253 100644
  		Timeouts: &schema.ResourceTimeout{
 -			Create: schema.DefaultTimeout(10 * time.Minute),
 -			Delete: schema.DefaultTimeout(15 * time.Minute),
-+			Create: schema.DefaultTimeout(10 * time.Second),
++			Create: schema.DefaultTimeout(20 * time.Second),
 +			Delete: schema.DefaultTimeout(15 * time.Second),
  		},
  
@@ -366,7 +366,7 @@ index e054f82987..08aeb6cf70 100644
  		Timeouts: &schema.ResourceTimeout{
 -			Create: schema.DefaultTimeout(10 * time.Minute),
 -			Delete: schema.DefaultTimeout(20 * time.Minute),
-+			Create: schema.DefaultTimeout(10 * time.Second),
++			Create: schema.DefaultTimeout(20 * time.Second),
 +			Delete: schema.DefaultTimeout(20 * time.Second),
  		},
  

tests/terraformtests/terraform-tests.success.txt

@@ -40,6 +40,8 @@ ec2:
   - TestAccEC2InternetGateway_
   - TestAccEC2NATGateway_
   - TestAccEC2RouteTableAssociation_
+  - TestAccEC2SpotInstanceRequest_disappears
+  - TestAccEC2SpotInstanceRequest_interruptUpdate
   - TestAccEC2VPCEndpointService_
   - TestAccEC2VPNGateway_
   - TestAccEC2VPNGatewayAttachment_

tests/test_ec2/test_spot_instances.py

@@ -7,7 +7,6 @@ import pytz
 import sure  # noqa # pylint: disable=unused-import
 
 from moto import mock_ec2, settings
-from moto.ec2.models import ec2_backends
 from moto.core.utils import iso_8601_datetime_with_milliseconds
 from tests import EXAMPLE_AMI_ID
 from uuid import uuid4
@@ -90,8 +89,8 @@ def test_request_spot_instances():
     requests.should.have.length_of(1)
     request = requests[0]
 
-    request["State"].should.equal("open")
+    request["State"].should.equal("active")
-    request["SpotPrice"].should.equal("0.5")
+    request["SpotPrice"].should.equal("0.500000")
     request["Type"].should.equal("one-time")
     request["ValidFrom"].should.equal(start_dt)
     request["ValidUntil"].should.equal(end_dt)
@@ -129,13 +128,14 @@ def test_request_spot_instances_default_arguments():
     requests.should.have.length_of(1)
     request = requests[0]
 
-    request["State"].should.equal("open")
+    request["State"].should.equal("active")
-    request["SpotPrice"].should.equal("0.5")
+    request["SpotPrice"].should.equal("0.500000")
     request["Type"].should.equal("one-time")
     request.shouldnt.contain("ValidFrom")
     request.shouldnt.contain("ValidUntil")
     request.shouldnt.contain("LaunchGroup")
     request.shouldnt.contain("AvailabilityZoneGroup")
+    request.should.have.key("InstanceInterruptionBehavior").equals("terminate")
 
     launch_spec = request["LaunchSpecification"]
 
@@ -153,7 +153,7 @@ def test_request_spot_instances_default_arguments():
 
 
 @mock_ec2
-def test_cancel_spot_instance_request_boto3():
+def test_cancel_spot_instance_request():
     client = boto3.client("ec2", region_name="us-west-1")
 
     rsi = client.request_spot_instances(
@@ -169,7 +169,7 @@ def test_cancel_spot_instance_request_boto3():
     request.should.have.key("CreateTime")
     request.should.have.key("Type").equal("one-time")
     request.should.have.key("SpotInstanceRequestId")
-    request.should.have.key("SpotPrice").equal("0.5")
+    request.should.have.key("SpotPrice").equal("0.500000")
     request["LaunchSpecification"]["ImageId"].should.equal(EXAMPLE_AMI_ID)
 
     with pytest.raises(ClientError) as ex:
@@ -193,7 +193,7 @@ def test_cancel_spot_instance_request_boto3():
 
 
 @mock_ec2
-def test_request_spot_instances_fulfilled_boto3():
+def test_request_spot_instances_fulfilled():
     """
     Test that moto correctly fullfills a spot instance request
     """
@@ -210,22 +210,11 @@ def test_request_spot_instances_fulfilled_boto3():
     requests.should.have.length_of(1)
     request = requests[0]
 
-    request["State"].should.equal("open")
-
-    if not settings.TEST_SERVER_MODE:
-        ec2_backends["us-east-1"].spot_instance_requests[request_id].state = "active"
-
-        requests = client.describe_spot_instance_requests(
-            SpotInstanceRequestIds=[request_id]
-        )["SpotInstanceRequests"]
-        requests.should.have.length_of(1)
-        request = requests[0]
-
     request["State"].should.equal("active")
 
 
 @mock_ec2
-def test_tag_spot_instance_request_boto3():
+def test_tag_spot_instance_request():
     """
     Test that moto correctly tags a spot instance request
     """
@@ -252,7 +241,7 @@ def test_tag_spot_instance_request_boto3():
 
 
 @mock_ec2
-def test_get_all_spot_instance_requests_filtering_boto3():
+def test_get_all_spot_instance_requests_filtering():
     """
     Test that moto correctly filters spot instance requests
     """
@@ -280,14 +269,14 @@ def test_get_all_spot_instance_requests_filtering_boto3():
     )
 
     requests = client.describe_spot_instance_requests(
-        Filters=[{"Name": "state", "Values": ["active"]}]
+        Filters=[{"Name": "state", "Values": ["failed"]}]
     )["SpotInstanceRequests"]
     r_ids = [r["SpotInstanceRequestId"] for r in requests]
     r_ids.shouldnt.contain(request1_id)
     r_ids.shouldnt.contain(request2_id)
 
     requests = client.describe_spot_instance_requests(
-        Filters=[{"Name": "state", "Values": ["open"]}]
+        Filters=[{"Name": "state", "Values": ["active"]}]
     )["SpotInstanceRequests"]
     r_ids = [r["SpotInstanceRequestId"] for r in requests]
     r_ids.should.contain(request1_id)
@@ -322,6 +311,27 @@ def test_request_spot_instances_instance_lifecycle():
     instance["InstanceLifecycle"].should.equal("spot")
 
 
+@mock_ec2
+def test_request_spot_instances_with_tags():
+    client = boto3.client("ec2", region_name="us-east-1")
+    request = client.request_spot_instances(
+        SpotPrice="0.5",
+        TagSpecifications=[
+            {
+                "ResourceType": "spot-instances-request",
+                "Tags": [{"Key": "k", "Value": "v"}],
+            }
+        ],
+    )
+
+    request_id = request["SpotInstanceRequests"][0]["SpotInstanceRequestId"]
+
+    request = client.describe_spot_instance_requests(
+        SpotInstanceRequestIds=[request_id]
+    )["SpotInstanceRequests"][0]
+    request.should.have.key("Tags").equals([{"Key": "k", "Value": "v"}])
+
+
 @mock_ec2
 def test_launch_spot_instance_instance_lifecycle():
     client = boto3.client("ec2", region_name="us-east-1")
@@ -391,18 +401,20 @@ def test_spot_price_history():
 
 
 @mock_ec2
-def test_request_spot_instances_setting_instance_id_boto3():
+def test_request_spot_instances__instance_should_exist():
     client = boto3.client("ec2", region_name="us-east-1")
     request = client.request_spot_instances(
         SpotPrice="0.5", LaunchSpecification={"ImageId": EXAMPLE_AMI_ID}
     )
     request_id = request["SpotInstanceRequests"][0]["SpotInstanceRequestId"]
 
-    if not settings.TEST_SERVER_MODE:
+    request = client.describe_spot_instance_requests(
-        req = ec2_backends["us-east-1"].spot_instance_requests[request_id]
+        SpotInstanceRequestIds=[request_id]
-        req.state = "active"
+    )["SpotInstanceRequests"][0]
-        req.instance_id = "i-12345678"
+    request.should.have.key("InstanceId")
+    instance_id = request["InstanceId"]
 
-        request = client.describe_spot_instance_requests()["SpotInstanceRequests"][0]
+    response = client.describe_instances(InstanceIds=[instance_id])
-        assert request["State"] == "active"
+    instance = response["Reservations"][0]["Instances"][0]
-        assert request["InstanceId"] == "i-12345678"
+    instance.should.have.key("InstanceId").equals(instance_id)
+    instance.should.have.key("ImageId").equals(EXAMPLE_AMI_ID)