Improve AWS driver performance

The following adjustments improve performance at large scale:

* Save the quota object earlier in the state machine.  This is
  still using cached data, but it's not necessary to re-run this
  each time.

* Flatten iterators returned from cached methods.  Some of our
  methods intended to cache heavy list responses were in fact
  only caching the iterator, and re-iterating would end up
  re-running the request.  This change does two things: it causes
  the iteration to happen within the rate limit calculator so we
  have a better idea of how long it actually took, and it causes
  the actual data to be put in the cache so that we don't re-run
  the request.

* Don't create a second instance object after creating the
  instance.  The create call returns an instance object in the
  form that we expect already.  Avoid creating a second one which
  incurs another DescribeInstances call.

Change-Id: I73bc099b450879917ab923fb7371f8006b113d68
This commit is contained in:
James E. Blair 2022-05-04 11:02:08 -07:00
parent 57fba40079
commit d2ac0de07b
2 changed files with 12 additions and 12 deletions

View File

@ -128,11 +128,11 @@ class AwsCreateStateMachine(statemachine.StateMachine):
self.instance = self.adapter._createInstance(
self.label, self.image_external_id,
self.tags, self.hostname, self.log)
self.quota = self.adapter._getQuotaForInstanceType(
self.instance.instance_type)
self.state = self.INSTANCE_CREATING
if self.state == self.INSTANCE_CREATING:
self.quota = self.adapter._getQuotaForInstanceType(
self.instance.instance_type)
self.instance = self.adapter._refresh(self.instance)
if self.instance.state["Name"].lower() == "running":
@ -448,24 +448,24 @@ class AwsAdapter(statemachine.Adapter):
def _listInstances(self):
with self.non_mutating_rate_limiter(
self.log.debug, "Listed instances"):
return self.ec2.instances.all()
return list(self.ec2.instances.all())
@cachetools.func.ttl_cache(maxsize=1, ttl=10)
def _listVolumes(self):
with self.non_mutating_rate_limiter:
return self.ec2.volumes.all()
return list(self.ec2.volumes.all())
@cachetools.func.ttl_cache(maxsize=1, ttl=10)
def _listAmis(self):
# Note: this is overridden in tests due to the filter
with self.non_mutating_rate_limiter:
return self.ec2.images.filter(Owners=['self'])
return list(self.ec2.images.filter(Owners=['self']))
@cachetools.func.ttl_cache(maxsize=1, ttl=10)
def _listSnapshots(self):
# Note: this is overridden in tests due to the filter
with self.non_mutating_rate_limiter:
return self.ec2.snapshots.filter(OwnerIds=['self'])
return list(self.ec2.snapshots.filter(OwnerIds=['self']))
@cachetools.func.ttl_cache(maxsize=1, ttl=10)
def _listObjects(self):
@ -475,7 +475,7 @@ class AwsAdapter(statemachine.Adapter):
bucket = self.s3.Bucket(bucket_name)
with self.non_mutating_rate_limiter:
return bucket.objects.all()
return list(bucket.objects.all())
def _getLatestImageIdByFilters(self, image_filters):
# Normally we would decorate this method, but our cache key is
@ -487,9 +487,9 @@ class AwsAdapter(statemachine.Adapter):
return val
with self.non_mutating_rate_limiter:
res = self.ec2_client.describe_images(
res = list(self.ec2_client.describe_images(
Filters=image_filters
).get("Images")
).get("Images"))
images = sorted(
res,
@ -598,7 +598,7 @@ class AwsAdapter(statemachine.Adapter):
log.debug(f"Creating VM {hostname}")
instances = self.ec2.create_instances(**args)
log.debug(f"Created VM {hostname} as instance {instances[0].id}")
return self.ec2.Instance(instances[0].id)
return instances[0]
def _deleteInstance(self, external_id, log=None):
if log is None:

View File

@ -128,7 +128,7 @@ class FakeAws:
raise NotImplementedError()
def _listAmis(self):
return self.ec2.images.filter()
return list(self.ec2.images.filter())
def _listSnapshots(self):
return self.ec2.snapshots.filter()
return list(self.ec2.snapshots.filter())