Add locking so that cASO does not run in parallel

If cASO runs in parallel it will end up providing duplicate records. We
add now a locking mechanism so that it does not run in parallel, but it
locks until the first execution has ended.

Closes #34

caso/manager.py

@@ -15,10 +15,12 @@
 # under the License.
 
 import datetime
+import os
 import os.path
 
 import dateutil.parser
 from dateutil import tz
+from oslo_concurrency import lockutils
 from oslo_config import cfg
 
 import caso.extract.manager
@@ -37,6 +39,16 @@
                help='Spool directory.'),
 ]
 
+override_lock = cfg.StrOpt(
+    "lock_path",
+    default=os.environ.get("CASO_LOCK_PATH", "$spooldir"),
+    help="Directory to use for lock files. For security, the specified "
+         "directory should only be writable by the user running the "
+         "processes that need locking. Defaults to environment variable "
+         "CASO_LOCK_PATH or $spooldir"
+)
+opts.append(override_lock)
+
 cli_opts = [
     cfg.BoolOpt('dry-run',
                 deprecated_name='dry_run',
@@ -50,6 +62,8 @@
 CONF.register_opts(opts)
 CONF.register_cli_opts(cli_opts)
 
+CONF.set_override("lock_path", override_lock, group="oslo_concurrency")
+
 
 class Manager(object):
     def __init__(self):
@@ -74,6 +88,7 @@ def lastrun(self):
             raise
         return date
 
+    @lockutils.synchronized("caso_should_not_run_in_parallel", external=True)
     def run(self):
         records = self.extractor_manager.get_records(lastrun=self.lastrun)
         if not CONF.dry_run:

caso/tests/test_manager.py

@@ -20,6 +20,7 @@
 
 from dateutil import tz
 import mock
+from oslo_concurrency.fixture import lockutils as lock_fixture
 import six
 
 from caso import manager
@@ -28,6 +29,7 @@
 
 class TestCasoManager(base.TestCase):
     def setUp(self):
+        self.useFixture(lock_fixture.ExternalLockFixture())
         super(TestCasoManager, self).setUp()
         self.patchers = {
             "makedirs": mock.patch('caso.utils.makedirs'),

etc/caso/caso-config-generator.conf

@@ -4,3 +4,4 @@ wrap_width = 79
 namespace = caso
 namespace = caso.config
 namespace = oslo.log
+namespace = oslo.concurrency

etc/caso/caso.conf.sample

@@ -13,6 +13,11 @@
 # Spool directory. (string value)
 #spooldir = /var/spool/caso
 
+# Directory to use for lock files. For security, the specified directory should
+# only be writable by the user running the processes that need locking.
+# Defaults to environment variable CASO_LOCK_PATH or $spooldir (string value)
+#lock_path = $spooldir
+
 # Extract records but do not push records to SSM. This will not update the last
 # run date. (boolean value)
 # Deprecated group/name - [DEFAULT]/dry_run
@@ -281,6 +286,22 @@
 #port = 5000
 
 
+[oslo_concurrency]
+
+#
+# From oslo.concurrency
+#
+
+# Enables or disables inter-process locks. (boolean value)
+#disable_process_locking = false
+
+# Directory to use for lock files.  For security, the specified directory
+# should only be writable by the user running the processes that need locking.
+# Defaults to environment variable OSLO_LOCK_PATH. If external locks are used,
+# a lock path must be set. (string value)
+#lock_path = <None>
+
+
 [ssm]
 
 #

requirements.txt

@@ -3,6 +3,7 @@
 # process, which may cause wedges in the gate later.
 pbr>=1.6
 oslo.config>=2.3.0 # Apache-2.0
+oslo.concurrency
 oslo.log>=1.8.0 # Apache-2.0
 oslo.utils!=2.6.0,>=2.0.0 # Apache-2.0