A resource can be imported by setting the `import` property in the resource options bag when instantiating a resource. In order to successfully import a resource, its desired configuration (i.e. its inputs) must not differ from its actual configuration (i.e. its state) as calculated by the resource's provider. There are a few interesting state transitions hiding here when importing a resource: 1. No prior resource exists in the checkpoint file. In this case, the resource is simply imported. 2. An external resource exists in the checkpoint file. In this case, the resource is imported and the old external state is discarded. 3. A non-external resource exists in the checkpoint file and its ID is different from the ID to import. In this case, the new resource is imported and the old resource is deleted. 4. A non-external resource exists in the checkpoint file, but the ID is the same as the ID to import. In this case, the import ID is ignored and the resource is treated as it would be in all cases except for changes that would replace the resource. In that case, the step generator issues an error that indicates that the import ID should be removed: were we to move forward with the replace, the new state of the stack would fall under case (3), which is almost certainly not what the user intends. Fixes #1662.
350 lines
14 KiB
Python
350 lines
14 KiB
Python
# Copyright 2016-2018, Pulumi Corporation.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
"""
|
|
Utility functions and classes for testing the Python language host.
|
|
"""
|
|
|
|
import asyncio
|
|
import unittest
|
|
from collections import namedtuple
|
|
from concurrent import futures
|
|
from inspect import signature
|
|
import logging
|
|
import subprocess
|
|
from os import path
|
|
import grpc
|
|
from pulumi.runtime import proto, rpc
|
|
from pulumi.runtime.proto import resource_pb2_grpc, language_pb2_grpc, engine_pb2_grpc, engine_pb2, provider_pb2
|
|
from google.protobuf import empty_pb2, struct_pb2
|
|
|
|
# gRPC by default logs exceptions to the root `logging` logger. We don't
|
|
# want this because it spews garbage to stderr and messes up our beautiful
|
|
# test output. Just turn it off.
|
|
logging.disable(level=logging.CRITICAL)
|
|
|
|
|
|
class LanghostMockResourceMonitor(proto.ResourceMonitorServicer):
|
|
"""
|
|
Implementation of proto.ResourceMonitorServicer for use in tests. Tests themselves
|
|
should not have to interact with this class.
|
|
|
|
This class encapsulates all gRPC details so that test authors only have to override
|
|
the `invoke`, `read_resource`, `register_resource`, and `register_resource_outputs`
|
|
functions on `LanghostTest` in order to get custom behavior.
|
|
"""
|
|
|
|
def __init__(self, langhost_test, dryrun):
|
|
self.reg_count = 0
|
|
self.registrations = {}
|
|
self.dryrun = dryrun
|
|
self.langhost_test = langhost_test
|
|
|
|
def Invoke(self, request, context):
|
|
args = rpc.deserialize_properties(request.args)
|
|
failures, ret = self.langhost_test.invoke(context, request.tok, args, request.provider, request.version)
|
|
failures_rpc = list(map(
|
|
lambda fail: provider_pb2.CheckFailure(property=fail["property"], reason=fail["reason"]), failures))
|
|
|
|
loop = asyncio.new_event_loop()
|
|
ret_proto = loop.run_until_complete(rpc.serialize_properties(ret, {}))
|
|
loop.close()
|
|
fields = {"failures": failures_rpc, "return": ret_proto}
|
|
return proto.InvokeResponse(**fields)
|
|
|
|
def ReadResource(self, request, context):
|
|
type_ = request.type
|
|
name = request.name
|
|
id_ = request.id
|
|
parent = request.parent
|
|
state = rpc.deserialize_properties(request.properties)
|
|
dependencies = sorted(list(request.dependencies))
|
|
provider = request.provider
|
|
version = request.version
|
|
outs = self.langhost_test.read_resource(context, type_, name, id_,
|
|
parent, state, dependencies, provider, version)
|
|
if "properties" in outs:
|
|
loop = asyncio.new_event_loop()
|
|
props_proto = loop.run_until_complete(rpc.serialize_properties(outs["properties"], {}))
|
|
loop.close()
|
|
else:
|
|
props_proto = None
|
|
return proto.ReadResourceResponse(
|
|
urn=outs.get("urn"), properties=props_proto)
|
|
|
|
def RegisterResource(self, request, context):
|
|
type_ = request.type
|
|
name = request.name
|
|
props = rpc.deserialize_properties(request.object)
|
|
deps = sorted(list(request.dependencies))
|
|
parent = request.parent
|
|
custom = request.custom
|
|
protect = request.protect
|
|
provider = request.provider
|
|
delete_before_replace = request.deleteBeforeReplace
|
|
ignore_changes = sorted(list(request.ignoreChanges))
|
|
version = request.version
|
|
import_ = request.importId
|
|
|
|
property_dependencies = {}
|
|
for key, value in request.propertyDependencies.items():
|
|
property_dependencies[key] = sorted(list(value.urns))
|
|
|
|
outs = {}
|
|
if type_ != "pulumi:pulumi:Stack":
|
|
rrsig = signature(self.langhost_test.register_resource)
|
|
args = [context, self.dryrun, type_, name, props, deps, parent, custom, protect, provider, property_dependencies, delete_before_replace, ignore_changes, version, import_]
|
|
outs = self.langhost_test.register_resource(*args[0:len(rrsig.parameters)])
|
|
if outs.get("urn"):
|
|
urn = outs["urn"]
|
|
self.registrations[urn] = {
|
|
"type": type_,
|
|
"name": name,
|
|
"props": props
|
|
}
|
|
|
|
self.reg_count += 1
|
|
else:
|
|
# Record the Stack's registration so that it can be the target of register_resource_outputs
|
|
# later on.
|
|
urn = self.langhost_test.make_urn(type_, "teststack")
|
|
self.registrations[urn] = {
|
|
"type": type_,
|
|
"name": "somestack",
|
|
"props": {}
|
|
}
|
|
|
|
return proto.RegisterResourceResponse(urn=urn, id="teststack", object=None)
|
|
if "object" in outs:
|
|
loop = asyncio.new_event_loop()
|
|
obj_proto = loop.run_until_complete(rpc.serialize_properties(outs["object"], {}))
|
|
loop.close()
|
|
else:
|
|
obj_proto = None
|
|
return proto.RegisterResourceResponse(
|
|
urn=outs.get("urn"), id=outs.get("id"), object=obj_proto)
|
|
|
|
def RegisterResourceOutputs(self, request, context):
|
|
urn = request.urn
|
|
outs = rpc.deserialize_properties(request.outputs)
|
|
res = self.registrations.get(urn)
|
|
if res:
|
|
self.langhost_test.register_resource_outputs(
|
|
context, self.dryrun, urn, res["type"], res["name"], res["props"], outs)
|
|
return empty_pb2.Empty()
|
|
|
|
|
|
class MockEngine(proto.EngineServicer):
|
|
"""
|
|
Implementation of the proto.EngineServicer protocol for use in tests. Like the
|
|
above class, we encapsulate all gRPC details here so that test writers only have
|
|
to override methods on LanghostTest.
|
|
"""
|
|
def Log(self, request, context):
|
|
if request.severity == engine_pb2.ERROR:
|
|
print(f"error: {request.message}")
|
|
return empty_pb2.Empty()
|
|
|
|
|
|
ResourceMonitorEndpoint = namedtuple('ResourceMonitorEndpoint',
|
|
['monitor', 'server', 'port'])
|
|
LanguageHostEndpoint = namedtuple('LanguageHostEndpoint', ['process', 'port'])
|
|
|
|
|
|
class LanghostTest(unittest.TestCase):
|
|
"""
|
|
Base class of all end-to-end tests of the Python language host.
|
|
|
|
The `run_test` method on this class executes a test by mocking the
|
|
Pulumi Engine gRPC endpoint and running the language host exactly the same
|
|
way that the Pulumi CLI will. Once the program completes and the language host
|
|
exits, `run_test` will assert the expected errors and/or resource registration
|
|
counts specified by the individual test.
|
|
|
|
Check out README.md in this directory for more details.
|
|
"""
|
|
|
|
def run_test(self,
|
|
project=None,
|
|
stack=None,
|
|
program=None,
|
|
pwd=None,
|
|
args=None,
|
|
config=None,
|
|
expected_resource_count=None,
|
|
expected_error=None,
|
|
expected_stderr_contains=None):
|
|
"""
|
|
Runs a language host test. The basic flow of a language host test is that
|
|
a test is launched using the real language host while mocking out the resource
|
|
monitor RPC interface and, once the program completes, assertions are made about
|
|
the sorts of resource monitor calls that were made.
|
|
|
|
:param project: The name of the project in which the program will run.
|
|
:param stack: The name of the stack in which the program will run.
|
|
:param program: The path to the program the langhost should execute.
|
|
:param pwd: The working directory the langhost should use when running the program.
|
|
:param args: Arguments to the program.
|
|
:param config: Configuration keys for the program.
|
|
:param expected_resource_count: The number of resources this program is expected to create.
|
|
:param expected_error: If present, the expected error that should arise when running this program.
|
|
:param expected_stderr_contains: If present, the standard error of the process should contain this string
|
|
"""
|
|
# For each test case, we'll do a preview followed by an update.
|
|
for dryrun in [True, False]:
|
|
# For these tests, we are the resource monitor. The `invoke`, `read_resource`,
|
|
# `register_resource`, and `register_resource_outputs` methods on this class
|
|
# will be used to implement the resource monitor gRPC interface. The tests
|
|
# that we spawn will connect to us directly.
|
|
monitor = self._create_mock_resource_monitor(dryrun)
|
|
|
|
# Now we'll launch the language host, which will in turn launch code that drives
|
|
# the test.
|
|
langhost = self._create_language_host(monitor.port)
|
|
|
|
# Run the program with the langhost we just launched.
|
|
with grpc.insecure_channel("localhost:%d" % langhost.port) as channel:
|
|
grpc.channel_ready_future(channel).result()
|
|
stub = language_pb2_grpc.LanguageRuntimeStub(channel)
|
|
result = self._run_program(stub, monitor, project, stack,
|
|
program, pwd, args, config, dryrun)
|
|
|
|
# Tear down the language host process we just spun up.
|
|
langhost.process.kill()
|
|
stdout, stderr = langhost.process.communicate()
|
|
if not expected_stderr_contains and (stdout or stderr):
|
|
print("PREVIEW:" if dryrun else "UPDATE:")
|
|
print("stdout:", stdout.decode('utf-8'))
|
|
print("stderr:", stderr.decode('utf-8'))
|
|
|
|
# If we expected an error, assert that we saw it. Otherwise assert
|
|
# that there wasn't an error.
|
|
expected = expected_error or ""
|
|
self.assertEqual(result, expected)
|
|
|
|
if expected_stderr_contains:
|
|
if expected_stderr_contains not in str(stderr):
|
|
print("stderr:", str(stderr))
|
|
self.fail("expected stderr to contain '" + expected_stderr_contains + "'")
|
|
|
|
if expected_resource_count is not None:
|
|
self.assertEqual(expected_resource_count,
|
|
monitor.monitor.reg_count)
|
|
|
|
monitor.server.stop(0)
|
|
|
|
def invoke(self, _ctx, _token, _args, _provider):
|
|
"""
|
|
Method corresponding to the `Invoke` resource monitor RPC call.
|
|
Override for custom behavior or assertions.
|
|
|
|
Returns a tuple of a list of errors that returned and the returned
|
|
object, if the call was successful.
|
|
"""
|
|
return ([], {})
|
|
|
|
def read_resource(self, _ctx, _type, _name, _id, _parent, _state):
|
|
"""
|
|
Method corresponding to the `ReadResource` resource monitor RPC call.
|
|
Override for custom behavior or assertions.
|
|
|
|
Returns a resource that was read from existing state.
|
|
"""
|
|
return {}
|
|
|
|
def register_resource(self, _ctx, _dry_run, _type, _name, _resource,
|
|
_dependencies, _parent, _custom, _provider, _property_deps, _delete_before_replace, _import):
|
|
"""
|
|
Method corresponding to the `RegisterResource` resource monitor RPC call.
|
|
Override for custom behavior or assertions.
|
|
|
|
Returns a resource that was just created.
|
|
"""
|
|
return {}
|
|
|
|
def register_resource_outputs(self, _ctx, _dry_run, _urn, _type, _name,
|
|
_resource, _outputs):
|
|
"""
|
|
Method corresponding to the `RegisterResourceOutputs` resource monitor RPC call.
|
|
Override for custom behavior or assertirons.
|
|
|
|
Returns None.
|
|
"""
|
|
pass
|
|
|
|
def make_urn(self, type_, name):
|
|
"""
|
|
Makes an URN from a given resource type and name.
|
|
"""
|
|
return "%s::%s" % (type_, name)
|
|
|
|
def base_path(self):
|
|
"""
|
|
Returns the base path for language host tests.
|
|
"""
|
|
return path.dirname(__file__)
|
|
|
|
def _create_mock_resource_monitor(self, dryrun):
|
|
monitor = LanghostMockResourceMonitor(self, dryrun)
|
|
engine = MockEngine()
|
|
server = grpc.server(futures.ThreadPoolExecutor(max_workers=4))
|
|
|
|
resource_pb2_grpc.add_ResourceMonitorServicer_to_server(monitor, server)
|
|
engine_pb2_grpc.add_EngineServicer_to_server(engine, server)
|
|
|
|
port = server.add_insecure_port(address="0.0.0.0:0")
|
|
server.start()
|
|
return ResourceMonitorEndpoint(monitor, server, port)
|
|
|
|
def _create_language_host(self, port):
|
|
exec_path = path.join(path.dirname(__file__), "..", "..", "..", "cmd", "pulumi-language-python-exec")
|
|
proc = subprocess.Popen(
|
|
["pulumi-language-python", "--use-executor", exec_path, "localhost:%d" % port],
|
|
stdout=subprocess.PIPE,
|
|
stderr=subprocess.PIPE)
|
|
# The first line of output is the port that the language host gRPC server is listening on.
|
|
first_line = proc.stdout.readline()
|
|
try:
|
|
return LanguageHostEndpoint(proc, int(first_line))
|
|
except ValueError:
|
|
proc.kill()
|
|
stdout, stderr = proc.communicate()
|
|
print("language host did not respond with port")
|
|
print("stdout: ")
|
|
print(stdout)
|
|
print("stderr: ")
|
|
print(stderr)
|
|
raise
|
|
|
|
def _run_program(self, stub, monitor, project, stack, program, pwd, args,
|
|
config, dryrun):
|
|
args = {}
|
|
args["monitor_address"] = "localhost:%d" % monitor.port
|
|
args["project"] = project or "project"
|
|
args["stack"] = stack or "stack"
|
|
args["program"] = program
|
|
if pwd:
|
|
args["pwd"] = pwd
|
|
|
|
if args:
|
|
args["args"] = args
|
|
|
|
if config:
|
|
args["config"] = config
|
|
|
|
args["dryRun"] = dryrun
|
|
request = proto.RunRequest(**args)
|
|
self.assertTrue(request.IsInitialized())
|
|
resp = stub.Run(request)
|
|
return resp.error
|