atst/tests/domain/authnid/test_crl.py
dandds 0b5acde4c4 Stream-parse CRLs for caching file locations.
AT-AT needs to maintain a key-value CRL cache where each key is the DER
byte-string of the issuer and the value is a dictionary of the CRL file
path and expiration. This way when it checks a client certificate, it
can load the correct CRL by comparing the issuers. This is preferable to
loading all of the CRLs in-memory. However, it still requires that AT-AT
load and parse each CRL when the application boots. Because of the size
of the CRLs and their parsed, in-memory size, this leads to the
application spiking to use nearly 900MB of memory (resting usage is
around 50MB).

This change introduces a small function to ad-hoc parse the CRL and
obtain the information in the CRL we need: the issuer and the
expiration. It does this by reading the CRL byte-by-byte until it
reaches the ASN1 sequence that corresponds to the issuer, and then looks
ahead to find the nextUpdate field (i.e., the expiration date). The
CRLCache class uses this function to build its cache and JSON-serializes
the cache to disk. If another AT-AT application process finds the
serialized version, it will load that copy instead of rebuilding it. It
also entails a change to the function signature for the init method of
CRLCache: now it expects the CRL directory as its second argument,
instead of a list of locations.

The Python script invoked by `script/sync-crls` will rebuild the
location cache each time it's run. This means that when the Kubernetes
CronJob for CRLs runs, it will refresh the cache each time. When a new
application container boots, it will get the refreshed cache.

This also adds a nightly CircleCI job to sync the CRLs and test that the
ad-hoc parsing function returns the same result as a proper parsing
using the Python cryptography library. This provides extra insurance
that the function is returning correct results on real data.
2019-11-04 08:36:03 -05:00

231 lines
7.4 KiB
Python

# Import installed packages
import pytest
import re
import os
import shutil
from cryptography.hazmat.backends import default_backend
from cryptography.hazmat.primitives.serialization import Encoding
from OpenSSL import crypto
from atst.domain.authnid.crl import (
CRLCache,
CRLRevocationException,
CRLInvalidException,
NoOpCRLCache,
)
from atst.domain.authnid.crl.util import (
scan_for_issuer_and_next_update,
build_crl_locations_cache,
serialize_crl_locations_cache,
load_crl_locations_cache,
CRLParseError,
JSON_CACHE,
)
from tests.mocks import FIXTURE_EMAIL_ADDRESS, DOD_CN
from tests.utils import FakeLogger, parse_for_issuer_and_next_update
class MockX509Store:
def __init__(self):
self.crls = []
self.certs = []
def add_crl(self, crl):
self.crls.append(crl)
def add_cert(self, cert):
self.certs.append(cert)
def set_flags(self, flag):
pass
def test_can_build_crl_list(crl_file, ca_key, ca_file, make_crl, tmpdir):
crl = make_crl(ca_key)
dir_ = os.path.dirname(crl_file)
serialize_crl_locations_cache(dir_)
cache = CRLCache(ca_file, dir_, store_class=MockX509Store)
issuer_der = crl.issuer.public_bytes(default_backend())
assert len(cache.crl_cache.keys()) == 1
assert issuer_der in cache.crl_cache
assert cache.crl_cache[issuer_der]["location"] == crl_file
assert cache.crl_cache[issuer_der]["expires"] == crl.next_update
def test_can_build_trusted_root_list(app):
location = "ssl/server-certs/ca-chain.pem"
cache = CRLCache(
location, app.config["CRL_STORAGE_CONTAINER"], store_class=MockX509Store
)
with open(location) as f:
content = f.read()
assert len(cache.certificate_authorities.keys()) == content.count("BEGIN CERT")
def test_crl_validation_on_login(
app,
client,
ca_key,
ca_file,
crl_file,
rsa_key,
make_x509,
make_crl,
serialize_pki_object_to_disk,
):
good_cert = make_x509(rsa_key(), signer_key=ca_key, cn="luke")
bad_cert = make_x509(rsa_key(), signer_key=ca_key, cn="darth")
crl = make_crl(ca_key, expired_serials=[bad_cert.serial_number])
serialize_pki_object_to_disk(crl, crl_file, encoding=Encoding.DER)
crl_dir = os.path.dirname(crl_file)
cache = CRLCache(ca_file, crl_dir)
assert cache.crl_check(good_cert.public_bytes(Encoding.PEM).decode())
with pytest.raises(CRLRevocationException):
cache.crl_check(bad_cert.public_bytes(Encoding.PEM).decode())
def test_can_dynamically_update_crls(
ca_key,
ca_file,
crl_file,
rsa_key,
make_x509,
make_crl,
serialize_pki_object_to_disk,
):
crl_dir = os.path.dirname(crl_file)
cache = CRLCache(ca_file, crl_dir)
client_cert = make_x509(rsa_key(), signer_key=ca_key, cn="chewbacca")
client_pem = client_cert.public_bytes(Encoding.PEM)
assert cache.crl_check(client_pem)
revoked_crl = make_crl(ca_key, expired_serials=[client_cert.serial_number])
# override the original CRL with one that revokes client_cert
serialize_pki_object_to_disk(revoked_crl, crl_file, encoding=Encoding.DER)
with pytest.raises(CRLRevocationException):
assert cache.crl_check(client_pem)
def test_throws_error_for_missing_issuer(app):
cache = CRLCache(
"ssl/server-certs/ca-chain.pem", app.config["CRL_STORAGE_CONTAINER"]
)
# this cert is self-signed, and so the application does not have a
# corresponding CRL for it
cert = open("tests/fixtures/{}.crt".format(FIXTURE_EMAIL_ADDRESS), "rb").read()
with pytest.raises(CRLInvalidException) as exc:
assert cache.crl_check(cert)
(message,) = exc.value.args
# objects that the issuer is missing
assert "issuer" in message
# names the issuer we were expecting to find a CRL for; same as the
# certificate subject in this case because the cert is self-signed
assert DOD_CN in message
def test_multistep_certificate_chain():
cache = CRLCache("tests/fixtures/chain/ca-chain.pem", "tests/fixtures/chain/")
cert = open("tests/fixtures/chain/client.crt", "rb").read()
assert cache.crl_check(cert)
def test_no_op_crl_cache_logs_common_name():
logger = FakeLogger()
cert = open("ssl/client-certs/atat.mil.crt", "rb").read()
cache = NoOpCRLCache(logger=logger)
assert cache.crl_check(cert)
assert "ART.GARFUNKEL.1234567890" in logger.messages[-1]
def test_expired_crl_raises_CRLInvalidException_with_failover_config_false(
app, ca_file, expired_crl_file, ca_key, make_x509, rsa_key
):
client_cert = make_x509(rsa_key(), signer_key=ca_key, cn="chewbacca")
client_pem = client_cert.public_bytes(Encoding.PEM)
crl_dir = os.path.dirname(expired_crl_file)
crl_cache = CRLCache(ca_file, crl_dir)
with pytest.raises(CRLInvalidException):
crl_cache.crl_check(client_pem)
def test_expired_crl_passes_with_failover_config_true(
ca_file, expired_crl_file, ca_key, make_x509, rsa_key, crl_failover_open_app
):
client_cert = make_x509(rsa_key(), signer_key=ca_key, cn="chewbacca")
client_pem = client_cert.public_bytes(Encoding.PEM)
crl_dir = os.path.dirname(expired_crl_file)
crl_cache = CRLCache(ca_file, crl_dir)
assert crl_cache.crl_check(client_pem)
def test_updates_expired_certs(
rsa_key, ca_file, expired_crl_file, crl_file, ca_key, make_x509
):
"""
Given a CRLCache object with an expired CRL and a function for updating the
CRLs, the CRLCache should run the update function before checking a
certificate that requires the expired CRL.
"""
client_cert = make_x509(rsa_key(), signer_key=ca_key, cn="chewbacca")
client_pem = client_cert.public_bytes(Encoding.PEM)
def _crl_update_func():
shutil.copyfile(crl_file, expired_crl_file)
crl_dir = os.path.dirname(expired_crl_file)
crl_cache = CRLCache(ca_file, crl_dir, crl_update_func=_crl_update_func)
crl_cache.crl_check(client_pem)
def test_scan_for_issuer_and_next_update(crl_file):
parsed = parse_for_issuer_and_next_update(crl_file)
scanned = scan_for_issuer_and_next_update(crl_file)
assert parsed == scanned
@pytest.fixture
def bad_crl(tmpdir):
bad_file = tmpdir.join("bad.crl")
with open(bad_file, "wb") as bad:
bad.write(b"definitely not a crl")
return bad_file
def test_scan_for_issuer_and_next_update_with_bad_data(bad_crl):
with pytest.raises(CRLParseError):
scan_for_issuer_and_next_update(bad_crl)
def test_build_crl_locations_cache(crl_file):
issuer_der, next_update = parse_for_issuer_and_next_update(crl_file)
cache = build_crl_locations_cache([crl_file])
assert cache == {issuer_der: {"location": crl_file, "expires": next_update}}
def test_build_crl_locations_cache_with_bad_data(crl_file, bad_crl):
logger = FakeLogger()
issuer_der, next_update = parse_for_issuer_and_next_update(crl_file)
cache = build_crl_locations_cache([crl_file, bad_crl], logger=logger)
assert cache == {issuer_der: {"location": crl_file, "expires": next_update}}
assert logger.messages
assert str(bad_crl) in logger.messages[0]
def test_serialize_crl_locations_cache(crl_file, bad_crl):
dir_ = os.path.dirname(crl_file)
serialize_crl_locations_cache(dir_)
assert os.path.isfile("{}/{}".format(dir_, JSON_CACHE))
def test_load_crl_locations_cache(crl_file, bad_crl):
dir_ = os.path.dirname(crl_file)
serialize_crl_locations_cache(dir_)
cache = load_crl_locations_cache(dir_)
assert isinstance(cache, dict)