def setUp(self):
super(XDCRNewBaseTest, self).setUp()
self.clusters = self.get_clusters()
self.task = self.get_task()
self.taskmgr = self.get_task_mgr()
for cluster in self.clusters:
cluster_util = ClusterUtils(cluster, self.taskmgr)
cluster_util.add_all_nodes_then_rebalance(cluster.servers[1:])
def __init__(self, cb_clusters, task, taskmgr):
self.__cb_clusters = cb_clusters
self.task = task
self.task_manager = taskmgr
for cluster in self.__cb_clusters:
cluster.cluster_util = ClusterUtils(cluster, self.task_manager)
cluster.bucket_util = BucketUtils(cluster.cluster_util, self.task)
self.input = TestInputSingleton.input
self.init_parameters()
self.create_buckets()
self.log = logger.get("test")
def get_cb_cluster_by_name(self, name):
"""Return couchbase cluster object for given name.
@return: CBCluster object
"""
for cluster in self.clusters:
if cluster.name == name:
cluster.cluster_util = ClusterUtils(cluster, self.task_manager)
cluster.bucket_util = BucketUtils(cluster,
cluster.cluster_util,
self.task)
return cluster
raise Exception(
"Couchbase Cluster with name: {0} not exist".format(name))
def fetch_cb_collect_logs(self):
log_path = TestInputSingleton.input.param("logs_folder", "/tmp")
is_single_node_server = len(self.servers) == 1
for _, cluster in self.cb_clusters.items():
rest = RestConnection(cluster.master)
nodes = rest.get_nodes()
# Creating cluster_util object to handle multi_cluster scenario
cluster_util = ClusterUtils(cluster, self.task_manager)
status = cluster_util.trigger_cb_collect_on_cluster(
rest, nodes, is_single_node_server)
if status is True:
cluster_util.wait_for_cb_collect_to_complete(rest)
cluster_util.copy_cb_collect_logs(rest, nodes, cluster,
log_path)
else:
self.log.error("API perform_cb_collect returned False")
def tearDownEverything(self):
if self.skip_setup_cleanup:
return
for cluster in self.__cb_clusters:
cluster_util = ClusterUtils(cluster, self.task_manager)
bucket_util = BucketUtils(cluster, cluster_util, self.task)
try:
if hasattr(self,
'skip_buckets_handle') and self.skip_buckets_handle:
return
test_failed = (hasattr(self, '_resultForDoCleanups') and
len(self._resultForDoCleanups.failures or
self._resultForDoCleanups.errors)) or \
(hasattr(self, '_exc_info') and \
self._exc_info()[1] is not None)
if test_failed and TestInputSingleton.input.param("stop-on-failure", False) \
or self.input.param("skip_cleanup", False):
self.log.warn("CLEANUP WAS SKIPPED")
else:
if test_failed:
# collect logs here because we have not shut things down
if TestInputSingleton.input.param(
"get-cbcollect-info", False):
self.fetch_cb_collect_logs()
if TestInputSingleton.input.param('get_trace', None):
for server in cluster.servers:
try:
shell = RemoteMachineShellConnection(
server)
output, _ = shell.execute_command(
"ps -aef|grep %s" %
TestInputSingleton.input.param(
'get_trace', None))
output = shell.execute_command(
"pstack %s" %
output[0].split()[1].strip())
self.infra_log.debug(output[0])
shell.disconnect()
except:
pass
else:
self.log.critical("Skipping get_trace !!")
rest = RestConnection(cluster.master)
alerts = rest.get_alerts()
if alerts is not None and len(alerts) != 0:
self.infra_log.warn("Alerts found: {0}".format(alerts))
self.log.debug("Cleaning up cluster")
cluster_util.cluster_cleanup(bucket_util)
except BaseException as e:
# kill memcached
traceback.print_exc()
self.log.warning("Killing memcached due to {0}".format(e))
cluster_util.kill_memcached()
# increase case_number to retry tearDown in setup for the next test
self.case_number += 1000
finally:
if not self.input.param("skip_cleanup", False):
cluster_util.reset_cluster()
# stop all existing task manager threads
if self.cleanup:
self.cleanup = False
else:
cluster_util.reset_env_variables()
self.infra_log.info("========== tasks in thread pool ==========")
self.task_manager.print_tasks_in_pool()
self.infra_log.info("==========================================")
if not self.tear_down_while_setup:
self.task_manager.shutdown_task_manager()
self.task.shutdown(force=True)
def setUp(self):
self.input = TestInputSingleton.input
# Framework specific parameters
self.log_level = self.input.param("log_level", "info").upper()
self.infra_log_level = self.input.param("infra_log_level",
"info").upper()
self.skip_setup_cleanup = self.input.param("skip_setup_cleanup", False)
self.tear_down_while_setup = self.input.param("tear_down_while_setup",
True)
self.test_timeout = self.input.param("test_timeout", 3600)
self.thread_to_use = self.input.param("threads_to_use", 10)
self.case_number = self.input.param("case_number", 0)
# End of framework parameters
# Cluster level info settings
self.log_info = self.input.param("log_info", None)
self.log_location = self.input.param("log_location", None)
self.stat_info = self.input.param("stat_info", None)
self.port = self.input.param("port", None)
self.port_info = self.input.param("port_info", None)
self.servers = self.input.servers
self.__cb_clusters = []
self.num_servers = self.input.param("servers", len(self.servers))
self.primary_index_created = False
self.index_quota_percent = self.input.param("index_quota_percent",
None)
self.gsi_type = self.input.param("gsi_type", 'plasma')
# CBAS setting
self.jre_path = self.input.param("jre_path", None)
# End of cluster info parameters
# Bucket specific params
self.bucket_type = self.input.param("bucket_type",
Bucket.bucket_type.MEMBASE)
self.bucket_size = self.input.param("bucket_size", None)
self.bucket_lww = self.input.param("lww", True)
self.standard_buckets = self.input.param("standard_buckets", 1)
if self.standard_buckets > 10:
self.bucket_util.change_max_buckets(self.standard_buckets)
self.vbuckets = self.input.param("vbuckets", 1024)
self.num_replicas = self.input.param("replicas", 1)
self.active_resident_threshold = int(
self.input.param("active_resident_threshold", 100))
self.compression_mode = self.input.param("compression_mode", 'passive')
# End of bucket parameters
# Doc specific params
self.key_size = self.input.param("key_size", 0)
self.doc_size = self.input.param("doc_size", 10)
self.sub_doc_size = self.input.param("sub_doc_size", 10)
self.doc_type = self.input.param("doc_type", "json")
self.num_items = self.input.param("num_items", 100000)
self.target_vbucket = self.input.param("target_vbucket", None)
self.maxttl = self.input.param("maxttl", 0)
# End of doc specific parameters
# Transactions parameters
self.transaction_timeout = self.input.param("transaction_timeout", 100)
self.transaction_commit = self.input.param("transaction_commit", True)
self.update_count = self.input.param("update_count", 1)
self.sync = self.input.param("sync", True)
self.default_bucket = self.input.param("default_bucket", True)
self.num_buckets = self.input.param("num_buckets", 0)
self.atomicity = self.input.param("atomicity", False)
# end of transaction parameters
# Client specific params
self.sdk_client_type = self.input.param("sdk_client_type", "java")
self.sdk_compression = self.input.param("sdk_compression", True)
self.replicate_to = self.input.param("replicate_to", 0)
self.persist_to = self.input.param("persist_to", 0)
self.sdk_retries = self.input.param("sdk_retries", 5)
self.sdk_timeout = self.input.param("sdk_timeout", 5)
self.durability_level = self.input.param("durability", "")
# Doc Loader Params
self.process_concurrency = self.input.param("process_concurrency", 8)
self.batch_size = self.input.param("batch_size", 20)
self.ryow = self.input.param("ryow", False)
self.check_persistence = self.input.param("check_persistence", False)
# End of client specific parameters
# initial number of items in the cluster
self.services_init = self.input.param("services_init", None)
self.nodes_init = self.input.param("nodes_init", 1)
self.nodes_in = self.input.param("nodes_in", 1)
self.nodes_out = self.input.param("nodes_out", 1)
self.services_in = self.input.param("services_in", None)
self.forceEject = self.input.param("forceEject", False)
self.wait_timeout = self.input.param("wait_timeout", 60)
self.dgm_run = self.input.param("dgm_run", False)
self.verify_unacked_bytes = self.input.param("verify_unacked_bytes",
False)
self.disabled_consistent_view = self.input.param(
"disabled_consistent_view", None)
self.rebalanceIndexWaitingDisabled = self.input.param(
"rebalanceIndexWaitingDisabled", None)
self.rebalanceIndexPausingDisabled = self.input.param(
"rebalanceIndexPausingDisabled", None)
self.maxParallelIndexers = self.input.param("maxParallelIndexers",
None)
self.maxParallelReplicaIndexers = self.input.param(
"maxParallelReplicaIndexers", None)
self.quota_percent = self.input.param("quota_percent", None)
if not hasattr(self, 'skip_buckets_handle'):
self.skip_buckets_handle = self.input.param(
"skip_buckets_handle", False)
# Initiate logging variables
self.log = logging.getLogger("test")
self.infra_log = logging.getLogger("infra")
# Configure loggers
self.log.setLevel(self.log_level)
self.infra_log.setLevel(self.infra_log_level)
# Support lib objects for testcase execution
self.task_manager = TaskManager(self.thread_to_use)
self.task = ServerTasks(self.task_manager)
# End of library object creation
self.cleanup = False
self.nonroot = False
self.test_failure = None
self.__log_setup_status("started")
if len(self.input.clusters) > 1:
# Multi cluster setup
counter = 1
for _, nodes in self.input.clusters.iteritems():
self.__cb_clusters.append(
CBCluster(name="C%s" % counter, servers=nodes))
counter += 1
else:
# Single cluster
self.cluster = CBCluster(servers=self.servers)
self.__cb_clusters.append(self.cluster)
self.cluster_util = ClusterUtils(self.cluster, self.task_manager)
self.bucket_util = BucketUtils(self.cluster, self.cluster_util,
self.task)
for cluster in self.__cb_clusters:
shell = RemoteMachineShellConnection(cluster.master)
self.os_info = shell.extract_remote_info().type.lower()
if self.os_info != 'windows':
if cluster.master.ssh_username != "root":
self.nonroot = True
shell.disconnect()
break
""" some tests need to bypass checking cb server at set up
to run installation """
self.skip_init_check_cbserver = self.input.param(
"skip_init_check_cbserver", False)
try:
if self.skip_setup_cleanup:
self.buckets = self.bucket_util.get_all_buckets()
return
if not self.skip_init_check_cbserver:
for cluster in self.__cb_clusters:
self.cb_version = None
if RestHelper(RestConnection(
cluster.master)).is_ns_server_running():
"""
Since every new couchbase version, there will be new
features that test code won't work on previous release.
So we need to get couchbase version to filter out
those tests.
"""
self.cb_version = RestConnection(
cluster.master).get_nodes_version()
else:
self.log.debug("couchbase server does not run yet")
# We stopped supporting TAP protocol since 3.x and 3.x support also has stopped
self.protocol = "dcp"
self.services_map = None
self.__log_setup_status("started")
for cluster in self.__cb_clusters:
if not self.skip_buckets_handle and not self.skip_init_check_cbserver:
self.log.debug("Cleaning up cluster")
cluster_util = ClusterUtils(cluster, self.task_manager)
bucket_util = BucketUtils(cluster, cluster_util, self.task)
cluster_util.cluster_cleanup(bucket_util)
# avoid any cluster operations in setup for new upgrade
# & upgradeXDCR tests
if str(self.__class__).find('newupgradetests') != -1 or \
str(self.__class__).find('upgradeXDCR') != -1 or \
str(self.__class__).find('Upgrade_EpTests') != -1 or \
hasattr(self, 'skip_buckets_handle') and \
self.skip_buckets_handle:
self.log.warning(
"any cluster operation in setup will be skipped")
self.primary_index_created = True
self.__log_setup_status("finished")
return
# avoid clean up if the previous test has been tear down
if self.case_number == 1 or self.case_number > 1000:
if self.case_number > 1000:
self.log.warn(
"TearDown for previous test failed. will retry..")
self.case_number -= 1000
self.cleanup = True
if not self.skip_init_check_cbserver:
self.tearDownEverything()
self.tear_down_while_setup = False
if not self.skip_init_check_cbserver:
for cluster in self.__cb_clusters:
self.log.info("Initializing cluster")
cluster_util = ClusterUtils(cluster, self.task_manager)
# self.cluster_util.reset_cluster()
master_services = cluster_util.get_services(
cluster.servers[:1], self.services_init, start_node=0)
if master_services is not None:
master_services = master_services[0].split(",")
self.quota = self._initialize_nodes(
self.task,
cluster,
self.disabled_consistent_view,
self.rebalanceIndexWaitingDisabled,
self.rebalanceIndexPausingDisabled,
self.maxParallelIndexers,
self.maxParallelReplicaIndexers,
self.port,
self.quota_percent,
services=master_services)
cluster_util.change_env_variables()
cluster_util.change_checkpoint_params()
#cluster_util.add_all_nodes_then_rebalance(cluster.servers[1:])
self.log.info("{0} initialized".format(cluster))
else:
self.quota = ""
for cluster in self.__cb_clusters:
cluster_util = ClusterUtils(cluster, self.task_manager)
if self.log_info:
cluster_util.change_log_info()
if self.log_location:
cluster_util.change_log_location()
if self.stat_info:
cluster_util.change_stat_info()
if self.port_info:
cluster_util.change_port_info()
if self.port:
self.port = str(self.port)
self.__log_setup_status("finished")
if not self.skip_init_check_cbserver:
self.__log("started")
self.sleep(5)
except Exception, e:
traceback.print_exc()
self.task.shutdown(force=True)
self.fail(e)
def setUp(self):
super(OnCloudBaseTest, self).setUp()
# Framework specific parameters
for server in self.input.servers:
server.hosted_on_cloud = True
# End of framework parameters
# Cluster level info settings
self.servers = list()
self.capella = self.input.capella
self.num_clusters = self.input.param("num_clusters", 1)
# Bucket specific params
# Note: Over riding bucket_eviction_policy from CouchbaseBaseTest
self.bucket_eviction_policy = \
self.input.param("bucket_eviction_policy",
Bucket.EvictionPolicy.FULL_EVICTION)
# End of bucket parameters
# Doc Loader Params (Extension from cb_basetest)
self.delete_docs_at_end = self.input.param("delete_doc_at_end", True)
# End of client specific parameters
self.wait_timeout = self.input.param("wait_timeout", 120)
self.use_https = self.input.param("use_https", True)
self.enforce_tls = self.input.param("enforce_tls", True)
self.ipv4_only = self.input.param("ipv4_only", False)
self.ipv6_only = self.input.param("ipv6_only", False)
self.multiple_ca = self.input.param("multiple_ca", False)
CbServer.use_https = True
trust_all_certs()
# initialise pod object
url = self.input.capella.get("pod")
self.pod = Pod("https://%s" % url)
self.tenant = Tenant(self.input.capella.get("tenant_id"),
self.input.capella.get("capella_user"),
self.input.capella.get("capella_pwd"),
self.input.capella.get("secret_key"),
self.input.capella.get("access_key"))
'''
Be careful while using this flag.
This is only and only for stand-alone tests.
During bugs reproductions, when a crash is seen
stop_server_on_crash will stop the server
so that we can collect data/logs/dumps at the right time
'''
self.stop_server_on_crash = self.input.param("stop_server_on_crash",
False)
self.collect_data = self.input.param("collect_data", False)
self.validate_system_event_logs = \
self.input.param("validate_sys_event_logs", False)
self.nonroot = False
self.crash_warning = self.input.param("crash_warning", False)
self.rest_username = \
TestInputSingleton.input.membase_settings.rest_username
self.rest_password = \
TestInputSingleton.input.membase_settings.rest_password
self.log_setup_status(self.__class__.__name__, "started")
self.cluster_name_format = "C%s"
default_cluster_index = cluster_index = 1
self.capella_cluster_config = CapellaAPI.get_cluster_config(
environment="hosted",
description="Amazing Cloud",
single_az=False,
provider=self.input.param("provider", AWS.__str__).lower(),
region=self.input.param("region", AWS.Region.US_WEST_2),
timezone=Cluster.Timezone.PT,
plan=Cluster.Plan.DEV_PRO,
cluster_name="taf_cluster")
services = self.input.param("services", "data")
for service_group in services.split("-"):
service_group = service_group.split(":")
min_nodes = 3 if "data" in service_group else 2
service_config = CapellaAPI.get_cluster_config_spec(
services=service_group,
count=max(min_nodes, self.nodes_init),
compute=self.input.param("compute",
AWS.ComputeNode.VCPU4_RAM16),
storage_type=self.input.param("type", AWS.StorageType.GP3),
storage_size_gb=self.input.param("size", AWS.StorageSize.MIN),
storage_iops=self.input.param("iops", AWS.StorageIOPS.MIN))
if self.capella_cluster_config["place"]["hosted"]["provider"] \
!= AWS.__str__:
service_config["storage"].pop("iops")
self.capella_cluster_config["servers"].append(service_config)
self.tenant.project_id = \
TestInputSingleton.input.capella.get("project", None)
if not self.tenant.project_id:
CapellaAPI.create_project(self.pod, self.tenant, "a_taf_run")
# Comma separated cluster_ids [Eg: 123-456-789,111-222-333,..]
cluster_ids = TestInputSingleton.input.capella \
.get("clusters", "")
if cluster_ids:
cluster_ids = cluster_ids.split(",")
self.__get_existing_cluster_details(cluster_ids)
else:
tasks = list()
for _ in range(self.num_clusters):
cluster_name = self.cluster_name_format % cluster_index
self.capella_cluster_config["clusterName"] = \
"a_%s_%s_%sGB_%s" % (
self.input.param("provider", "aws"),
self.input.param("compute", "m5.xlarge")
.replace(".", ""),
self.input.param("size", 50),
cluster_name)
self.log.info(self.capella_cluster_config)
deploy_task = DeployCloud(self.pod,
self.tenant,
cluster_name,
self.capella_cluster_config,
timeout=self.wait_timeout)
self.task_manager.add_new_task(deploy_task)
tasks.append(deploy_task)
cluster_index += 1
for task in tasks:
self.task_manager.get_task_result(task)
self.assertTrue(task.result, "Cluster deployment failed!")
CapellaAPI.create_db_user(self.pod, self.tenant,
task.cluster_id, self.rest_username,
self.rest_password)
self.__populate_cluster_info(task.cluster_id, task.servers,
task.srv, task.name,
self.capella_cluster_config)
# Initialize self.cluster with first available cluster as default
self.cluster = self.cb_clusters[self.cluster_name_format %
default_cluster_index]
self.servers = self.cluster.servers
self.cluster_util = ClusterUtils(self.task_manager)
self.bucket_util = BucketUtils(self.cluster_util, self.task)
for _, cluster in self.cb_clusters.items():
self.cluster_util.print_cluster_stats(cluster)
self.cluster.edition = "enterprise"
self.sleep(10)
class OnCloudBaseTest(CouchbaseBaseTest):
def setUp(self):
super(OnCloudBaseTest, self).setUp()
# Framework specific parameters
for server in self.input.servers:
server.hosted_on_cloud = True
# End of framework parameters
# Cluster level info settings
self.servers = list()
self.capella = self.input.capella
self.num_clusters = self.input.param("num_clusters", 1)
# Bucket specific params
# Note: Over riding bucket_eviction_policy from CouchbaseBaseTest
self.bucket_eviction_policy = \
self.input.param("bucket_eviction_policy",
Bucket.EvictionPolicy.FULL_EVICTION)
# End of bucket parameters
# Doc Loader Params (Extension from cb_basetest)
self.delete_docs_at_end = self.input.param("delete_doc_at_end", True)
# End of client specific parameters
self.wait_timeout = self.input.param("wait_timeout", 120)
self.use_https = self.input.param("use_https", True)
self.enforce_tls = self.input.param("enforce_tls", True)
self.ipv4_only = self.input.param("ipv4_only", False)
self.ipv6_only = self.input.param("ipv6_only", False)
self.multiple_ca = self.input.param("multiple_ca", False)
CbServer.use_https = True
trust_all_certs()
# initialise pod object
url = self.input.capella.get("pod")
self.pod = Pod("https://%s" % url)
self.tenant = Tenant(self.input.capella.get("tenant_id"),
self.input.capella.get("capella_user"),
self.input.capella.get("capella_pwd"),
self.input.capella.get("secret_key"),
self.input.capella.get("access_key"))
'''
Be careful while using this flag.
This is only and only for stand-alone tests.
During bugs reproductions, when a crash is seen
stop_server_on_crash will stop the server
so that we can collect data/logs/dumps at the right time
'''
self.stop_server_on_crash = self.input.param("stop_server_on_crash",
False)
self.collect_data = self.input.param("collect_data", False)
self.validate_system_event_logs = \
self.input.param("validate_sys_event_logs", False)
self.nonroot = False
self.crash_warning = self.input.param("crash_warning", False)
self.rest_username = \
TestInputSingleton.input.membase_settings.rest_username
self.rest_password = \
TestInputSingleton.input.membase_settings.rest_password
self.log_setup_status(self.__class__.__name__, "started")
self.cluster_name_format = "C%s"
default_cluster_index = cluster_index = 1
self.capella_cluster_config = CapellaAPI.get_cluster_config(
environment="hosted",
description="Amazing Cloud",
single_az=False,
provider=self.input.param("provider", AWS.__str__).lower(),
region=self.input.param("region", AWS.Region.US_WEST_2),
timezone=Cluster.Timezone.PT,
plan=Cluster.Plan.DEV_PRO,
cluster_name="taf_cluster")
services = self.input.param("services", "data")
for service_group in services.split("-"):
service_group = service_group.split(":")
min_nodes = 3 if "data" in service_group else 2
service_config = CapellaAPI.get_cluster_config_spec(
services=service_group,
count=max(min_nodes, self.nodes_init),
compute=self.input.param("compute",
AWS.ComputeNode.VCPU4_RAM16),
storage_type=self.input.param("type", AWS.StorageType.GP3),
storage_size_gb=self.input.param("size", AWS.StorageSize.MIN),
storage_iops=self.input.param("iops", AWS.StorageIOPS.MIN))
if self.capella_cluster_config["place"]["hosted"]["provider"] \
!= AWS.__str__:
service_config["storage"].pop("iops")
self.capella_cluster_config["servers"].append(service_config)
self.tenant.project_id = \
TestInputSingleton.input.capella.get("project", None)
if not self.tenant.project_id:
CapellaAPI.create_project(self.pod, self.tenant, "a_taf_run")
# Comma separated cluster_ids [Eg: 123-456-789,111-222-333,..]
cluster_ids = TestInputSingleton.input.capella \
.get("clusters", "")
if cluster_ids:
cluster_ids = cluster_ids.split(",")
self.__get_existing_cluster_details(cluster_ids)
else:
tasks = list()
for _ in range(self.num_clusters):
cluster_name = self.cluster_name_format % cluster_index
self.capella_cluster_config["clusterName"] = \
"a_%s_%s_%sGB_%s" % (
self.input.param("provider", "aws"),
self.input.param("compute", "m5.xlarge")
.replace(".", ""),
self.input.param("size", 50),
cluster_name)
self.log.info(self.capella_cluster_config)
deploy_task = DeployCloud(self.pod,
self.tenant,
cluster_name,
self.capella_cluster_config,
timeout=self.wait_timeout)
self.task_manager.add_new_task(deploy_task)
tasks.append(deploy_task)
cluster_index += 1
for task in tasks:
self.task_manager.get_task_result(task)
self.assertTrue(task.result, "Cluster deployment failed!")
CapellaAPI.create_db_user(self.pod, self.tenant,
task.cluster_id, self.rest_username,
self.rest_password)
self.__populate_cluster_info(task.cluster_id, task.servers,
task.srv, task.name,
self.capella_cluster_config)
# Initialize self.cluster with first available cluster as default
self.cluster = self.cb_clusters[self.cluster_name_format %
default_cluster_index]
self.servers = self.cluster.servers
self.cluster_util = ClusterUtils(self.task_manager)
self.bucket_util = BucketUtils(self.cluster_util, self.task)
for _, cluster in self.cb_clusters.items():
self.cluster_util.print_cluster_stats(cluster)
self.cluster.edition = "enterprise"
self.sleep(10)
def tearDown(self):
self.shutdown_task_manager()
if self.sdk_client_pool:
self.sdk_client_pool.shutdown()
if self.skip_teardown_cleanup:
return
for name, cluster in self.cb_clusters.items():
self.log.info("Destroying cluster: {}".format(name))
CapellaAPI.destroy_cluster(cluster)
CapellaAPI.delete_project(self.pod, self.tenant)
def __get_existing_cluster_details(self, cluster_ids):
cluster_index = 1
for cluster_id in cluster_ids:
cluster_name = self.cluster_name_format % cluster_index
self.log.info("Fetching cluster details for: %s" % cluster_id)
CapellaAPI.wait_until_done(self.pod, self.tenant, cluster_id,
"Cluster not healthy")
cluster_info = CapellaAPI.get_cluster_info(self.pod, self.tenant,
cluster_id)
cluster_srv = cluster_info.get("endpointsSrv")
CapellaAPI.allow_my_ip(self.pod, self.tenant, cluster_id)
CapellaAPI.create_db_user(self.pod, self.tenant, cluster_id,
self.rest_username, self.rest_password)
servers = CapellaAPI.get_nodes(self.pod, self.tenant, cluster_id)
self.__populate_cluster_info(cluster_id, servers, cluster_srv,
cluster_name, cluster_info)
self.__populate_cluster_buckets(self.cb_clusters[cluster_name])
def __populate_cluster_info(self, cluster_id, servers, cluster_srv,
cluster_name, service_config):
nodes = list()
for server in servers:
temp_server = TestInputServer()
temp_server.ip = server.get("hostname")
temp_server.hostname = server.get("hostname")
temp_server.services = server.get("services")
temp_server.port = "18091"
temp_server.rest_username = self.rest_username
temp_server.rest_password = self.rest_password
temp_server.hosted_on_cloud = True
temp_server.memcached_port = "11207"
nodes.append(temp_server)
cluster = CBCluster(username=self.rest_username,
password=self.rest_password,
servers=[None] * 40)
cluster.id = cluster_id
cluster.srv = cluster_srv
cluster.cluster_config = service_config
cluster.pod = self.pod
cluster.tenant = self.tenant
for temp_server in nodes:
if "Data" in temp_server.services:
cluster.kv_nodes.append(temp_server)
if "Query" in temp_server.services:
cluster.query_nodes.append(temp_server)
if "Index" in temp_server.services:
cluster.index_nodes.append(temp_server)
if "Eventing" in temp_server.services:
cluster.eventing_nodes.append(temp_server)
if "Analytics" in temp_server.services:
cluster.cbas_nodes.append(temp_server)
if "FTS" in temp_server.services:
cluster.fts_nodes.append(temp_server)
cluster.master = cluster.kv_nodes[0]
self.tenant.clusters.update({cluster.id: cluster})
cluster.master = cluster.kv_nodes[0]
self.tenant.clusters.update({cluster.id: cluster})
self.cb_clusters[cluster_name] = cluster
self.cb_clusters[cluster_name].cloud_cluster = True
def __populate_cluster_buckets(self, cluster):
self.log.debug("Fetching bucket details from cluster %s" % cluster.id)
buckets = json.loads(
CapellaAPI.get_all_buckets(cluster).content)["buckets"]["data"]
for bucket in buckets:
bucket = bucket["data"]
bucket_obj = Bucket({
Bucket.name:
bucket["name"],
Bucket.ramQuotaMB:
bucket["memoryAllocationInMb"],
Bucket.replicaNumber:
bucket["replicas"],
Bucket.conflictResolutionType:
bucket["bucketConflictResolution"],
Bucket.flushEnabled:
bucket["flush"],
Bucket.durabilityMinLevel:
bucket["durabilityLevel"],
Bucket.maxTTL:
bucket["timeToLive"],
})
bucket_obj.uuid = bucket["id"]
bucket_obj.stats.itemCount = bucket["stats"]["itemCount"]
bucket_obj.stats.memUsed = bucket["stats"]["memoryUsedInMib"]
cluster.buckets.append(bucket_obj)
def setUp(self):
self.input = TestInputSingleton.input
# Framework specific parameters
self.log_level = self.input.param("log_level", "info").upper()
self.infra_log_level = self.input.param("infra_log_level",
"error").upper()
self.skip_setup_cleanup = self.input.param("skip_setup_cleanup", False)
self.tear_down_while_setup = self.input.param("tear_down_while_setup",
True)
self.test_timeout = self.input.param("test_timeout", 3600)
self.thread_to_use = self.input.param("threads_to_use", 30)
self.case_number = self.input.param("case_number", 0)
# End of framework parameters
# Cluster level info settings
self.log_info = self.input.param("log_info", None)
self.log_location = self.input.param("log_location", None)
self.stat_info = self.input.param("stat_info", None)
self.port = self.input.param("port", None)
self.port_info = self.input.param("port_info", None)
self.servers = self.input.servers
self.cb_clusters = OrderedDict()
self.num_servers = self.input.param("servers", len(self.servers))
self.primary_index_created = False
self.index_quota_percent = self.input.param("index_quota_percent",
None)
self.gsi_type = self.input.param("gsi_type", 'plasma')
# CBAS setting
self.jre_path = self.input.param("jre_path", None)
self.enable_dp = self.input.param("enable_dp", False)
# End of cluster info parameters
# Bucket specific params
self.bucket_type = self.input.param("bucket_type", Bucket.Type.MEMBASE)
self.bucket_ttl = self.input.param("bucket_ttl", 0)
self.bucket_size = self.input.param("bucket_size", None)
self.bucket_conflict_resolution_type = \
self.input.param("bucket_conflict_resolution",
Bucket.ConflictResolution.SEQ_NO)
self.bucket_replica_index = self.input.param("bucket_replica_index", 1)
self.bucket_eviction_policy = \
self.input.param("bucket_eviction_policy",
Bucket.EvictionPolicy.VALUE_ONLY)
self.flush_enabled = self.input.param("flushEnabled",
Bucket.FlushBucket.DISABLED)
self.bucket_time_sync = self.input.param("bucket_time_sync", False)
self.standard_buckets = self.input.param("standard_buckets", 1)
self.num_replicas = self.input.param("replicas", Bucket.ReplicaNum.ONE)
self.active_resident_threshold = \
int(self.input.param("active_resident_threshold", 100))
self.compression_mode = \
self.input.param("compression_mode",
Bucket.CompressionMode.PASSIVE)
self.bucket_storage = \
self.input.param("bucket_storage",
Bucket.StorageBackend.couchstore)
if self.bucket_storage == Bucket.StorageBackend.magma:
self.bucket_eviction_policy = Bucket.EvictionPolicy.FULL_EVICTION
self.scope_name = self.input.param("scope", CbServer.default_scope)
self.collection_name = self.input.param("collection",
CbServer.default_collection)
self.bucket_durability_level = self.input.param(
"bucket_durability", Bucket.DurabilityLevel.NONE).upper()
self.bucket_purge_interval = self.input.param("bucket_purge_interval",
1)
self.bucket_durability_level = \
BucketDurability[self.bucket_durability_level]
# End of bucket parameters
# Doc specific params
self.key = self.input.param("key", "test_docs")
self.key_size = self.input.param("key_size", 8)
self.doc_size = self.input.param("doc_size", 256)
self.sub_doc_size = self.input.param("sub_doc_size", 10)
self.doc_type = self.input.param("doc_type", "json")
self.num_items = self.input.param("num_items", 100000)
self.target_vbucket = self.input.param("target_vbucket", None)
self.maxttl = self.input.param("maxttl", 0)
self.random_exp = self.input.param("random_exp", False)
self.randomize_doc_size = self.input.param("randomize_doc_size", False)
self.randomize_value = self.input.param("randomize_value", False)
self.rev_write = self.input.param("rev_write", False)
self.rev_read = self.input.param("rev_read", False)
self.rev_update = self.input.param("rev_update", False)
self.rev_del = self.input.param("rev_del", False)
self.random_key = self.input.param("random_key", False)
self.mix_key_size = self.input.param("mix_key_size", False)
# End of doc specific parameters
# Transactions parameters
self.transaction_timeout = self.input.param("transaction_timeout", 100)
self.transaction_commit = self.input.param("transaction_commit", True)
self.update_count = self.input.param("update_count", 1)
self.sync = self.input.param("sync", True)
self.default_bucket = self.input.param("default_bucket", True)
self.num_buckets = self.input.param("num_buckets", 0)
self.atomicity = self.input.param("atomicity", False)
self.defer = self.input.param("defer", False)
# end of transaction parameters
# Client specific params
self.sdk_client_type = self.input.param("sdk_client_type", "java")
self.replicate_to = self.input.param("replicate_to", 0)
self.persist_to = self.input.param("persist_to", 0)
self.sdk_retries = self.input.param("sdk_retries", 5)
self.sdk_timeout = self.input.param("sdk_timeout", 5)
self.time_unit = self.input.param("time_unit", "seconds")
self.durability_level = self.input.param("durability", "").upper()
self.sdk_client_pool = self.input.param("sdk_client_pool", None)
self.sdk_pool_capacity = self.input.param("sdk_pool_capacity", 1)
# Client compression settings
self.sdk_compression = self.input.param("sdk_compression", None)
compression_min_ratio = self.input.param("min_ratio", None)
compression_min_size = self.input.param("min_size", None)
if type(self.sdk_compression) is bool:
self.sdk_compression = {"enabled": self.sdk_compression}
if compression_min_size:
self.sdk_compression["minSize"] = compression_min_size
if compression_min_ratio:
self.sdk_compression["minRatio"] = compression_min_ratio
# Doc Loader Params
self.process_concurrency = self.input.param("process_concurrency", 20)
self.batch_size = self.input.param("batch_size", 2000)
self.dgm_batch = self.input.param("dgm_batch", 5000)
self.ryow = self.input.param("ryow", False)
self.check_persistence = self.input.param("check_persistence", False)
# End of client specific parameters
# initial number of items in the cluster
self.services_init = self.input.param("services_init", None)
self.nodes_init = self.input.param("nodes_init", 1)
self.nodes_in = self.input.param("nodes_in", 1)
self.nodes_out = self.input.param("nodes_out", 1)
self.services_in = self.input.param("services_in", None)
self.forceEject = self.input.param("forceEject", False)
self.wait_timeout = self.input.param("wait_timeout", 120)
self.verify_unacked_bytes = \
self.input.param("verify_unacked_bytes", False)
self.disabled_consistent_view = \
self.input.param("disabled_consistent_view", None)
self.rebalanceIndexWaitingDisabled = \
self.input.param("rebalanceIndexWaitingDisabled", None)
self.rebalanceIndexPausingDisabled = \
self.input.param("rebalanceIndexPausingDisabled", None)
self.maxParallelIndexers = \
self.input.param("maxParallelIndexers", None)
self.maxParallelReplicaIndexers = \
self.input.param("maxParallelReplicaIndexers", None)
self.quota_percent = self.input.param("quota_percent", 90)
self.skip_buckets_handle = self.input.param("skip_buckets_handle",
False)
# SDKClientPool object for creating generic clients across tasks
if self.sdk_client_pool is True:
self.init_sdk_pool_object()
# Initiate logging variables
self.log = logger.get("test")
self.infra_log = logger.get("infra")
self.cleanup_pcaps()
self.collect_pcaps = self.input.param("collect_pcaps", False)
if self.collect_pcaps:
self.start_collect_pcaps()
# variable for log collection using cbCollect
self.get_cbcollect_info = self.input.param("get-cbcollect-info", False)
# Variable for initializing the current (start of test) timestamp
self.start_timestamp = datetime.now()
'''
Be careful while using this flag.
This is only and only for stand-alone tests.
During bugs reproductions, when a crash is seen
stop_server_on_crash will stop the server
so that we can collect data/logs/dumps at the right time
'''
self.stop_server_on_crash = self.input.param("stop_server_on_crash",
False)
self.collect_data = self.input.param("collect_data", False)
# Configure loggers
self.log.setLevel(self.log_level)
self.infra_log.setLevel(self.infra_log_level)
# Support lib objects for testcase execution
self.task_manager = TaskManager(self.thread_to_use)
self.task = ServerTasks(self.task_manager)
# End of library object creation
self.sleep = sleep
self.cleanup = False
self.nonroot = False
self.test_failure = None
self.crash_warning = self.input.param("crash_warning", False)
self.summary = TestSummary(self.log)
# Populate memcached_port in case of cluster_run
cluster_run_base_port = ClusterRun.port
if int(self.input.servers[0].port) == ClusterRun.port:
for server in self.input.servers:
server.port = cluster_run_base_port
cluster_run_base_port += 1
# If not defined in node.ini under 'memcached_port' section
if server.memcached_port is CbServer.memcached_port:
server.memcached_port = \
ClusterRun.memcached_port \
+ (2 * (int(server.port) - ClusterRun.port))
self.log_setup_status(self.__class__.__name__, "started")
cluster_name_format = "C%s"
default_cluster_index = counter_index = 1
if len(self.input.clusters) > 1:
# Multi cluster setup
for _, nodes in self.input.clusters.iteritems():
cluster_name = cluster_name_format % counter_index
tem_cluster = CBCluster(name=cluster_name, servers=nodes)
self.cb_clusters[cluster_name] = tem_cluster
counter_index += 1
else:
# Single cluster
cluster_name = cluster_name_format % counter_index
self.cb_clusters[cluster_name] = CBCluster(name=cluster_name,
servers=self.servers)
# Initialize self.cluster with first available cluster as default
self.cluster = self.cb_clusters[cluster_name_format %
default_cluster_index]
self.cluster_util = ClusterUtils(self.cluster, self.task_manager)
self.bucket_util = BucketUtils(self.cluster_util, self.task)
if self.standard_buckets > 10:
self.bucket_util.change_max_buckets(self.cluster.master,
self.standard_buckets)
for cluster_name, cluster in self.cb_clusters.items():
shell = RemoteMachineShellConnection(cluster.master)
self.os_info = shell.extract_remote_info().type.lower()
if self.os_info != 'windows':
if cluster.master.ssh_username != "root":
self.nonroot = True
shell.disconnect()
break
shell.disconnect()
""" some tests need to bypass checking cb server at set up
to run installation """
self.skip_init_check_cbserver = \
self.input.param("skip_init_check_cbserver", False)
try:
if self.skip_setup_cleanup:
self.cluster.buckets = self.bucket_util.get_all_buckets(
self.cluster)
return
self.services_map = None
self.log_setup_status("BaseTestCase", "started")
for cluster_name, cluster in self.cb_clusters.items():
if not self.skip_buckets_handle \
and not self.skip_init_check_cbserver:
self.log.debug("Cleaning up cluster")
cluster_util = ClusterUtils(cluster, self.task_manager)
bucket_util = BucketUtils(cluster_util, self.task)
cluster_util.cluster_cleanup(bucket_util)
# Avoid cluster operations in setup for new upgrade / upgradeXDCR
if str(self.__class__).find('newupgradetests') != -1 or \
str(self.__class__).find('upgradeXDCR') != -1 or \
str(self.__class__).find('Upgrade_EpTests') != -1 or \
self.skip_buckets_handle:
self.log.warning("Cluster operation in setup will be skipped")
self.primary_index_created = True
self.log_setup_status("BaseTestCase", "finished")
return
# avoid clean up if the previous test has been tear down
if self.case_number == 1 or self.case_number > 1000:
if self.case_number > 1000:
self.log.warn("TearDown for prev test failed. Will retry")
self.case_number -= 1000
self.cleanup = True
if not self.skip_init_check_cbserver:
self.tearDownEverything()
self.tear_down_while_setup = False
if not self.skip_init_check_cbserver:
for cluster_name, cluster in self.cb_clusters.items():
self.log.info("Initializing cluster")
cluster_util = ClusterUtils(cluster, self.task_manager)
cluster_util.reset_cluster()
master_services = cluster_util.get_services(
cluster.servers[:1], self.services_init, start_node=0)
if master_services is not None:
master_services = master_services[0].split(",")
self.quota = self._initialize_nodes(
self.task,
cluster,
self.disabled_consistent_view,
self.rebalanceIndexWaitingDisabled,
self.rebalanceIndexPausingDisabled,
self.maxParallelIndexers,
self.maxParallelReplicaIndexers,
self.port,
self.quota_percent,
services=master_services)
cluster_util.change_env_variables()
cluster_util.change_checkpoint_params()
self.log.info("{0} initialized".format(cluster))
else:
self.quota = ""
# Enable dp_version since we need collections enabled
if self.enable_dp:
for server in self.cluster.servers:
shell_conn = RemoteMachineShellConnection(server)
cb_cli = CbCli(shell_conn)
cb_cli.enable_dp()
shell_conn.disconnect()
for cluster_name, cluster in self.cb_clusters.items():
cluster_util = ClusterUtils(cluster, self.task_manager)
if self.log_info:
cluster_util.change_log_info()
if self.log_location:
cluster_util.change_log_location()
if self.stat_info:
cluster_util.change_stat_info()
if self.port_info:
cluster_util.change_port_info()
if self.port:
self.port = str(self.port)
self.log_setup_status("BaseTestCase", "finished")
if not self.skip_init_check_cbserver:
self.__log("started")
except Exception as e:
traceback.print_exc()
self.task.shutdown(force=True)
self.fail(e)
def setUp(self):
super(EnterpriseBKRSNewBaseTest, self).setUp()
self.clusters = self.get_clusters()
self.master = self.servers[0]
self.task = self.get_task()
self.taskmgr = self.get_task_mgr()
self.backupset = Backupset()
self.cmd_ext = ""
self.should_fail = self.input.param("should-fail", False)
self.restore_should_fail = self.input.param("restore_should_fail",
False)
self.merge_should_fail = self.input.param("merge_should_fail", False)
self.database_path = COUCHBASE_DATA_PATH
cmd = 'curl -g {0}:8091/diag/eval -u {1}:{2} '.format(
self.master.ip, self.master.rest_username,
self.master.rest_password)
cmd += '-d "path_config:component_path(bin)."'
shell = RemoteMachineShellConnection(self.master)
output, error = shell.enable_diag_eval_on_non_local_hosts()
shell.disconnect()
bin_path = subprocess.check_output(cmd, shell=True)
if not self.skip_init_check_cbserver:
if "bin" not in bin_path:
self.fail("Check if cb server install on %s" % self.master.ip)
else:
self.cli_command_location = bin_path.replace('"', '') + "/"
self.debug_logs = self.input.param("debug-logs", False)
self.backupset.directory = self.input.param("dir", "/tmp/entbackup")
self.backupset.user_env = self.input.param("user-env", False)
self.backupset.passwd_env = self.input.param("passwd-env", False)
self.backupset.log_archive_env = self.input.param(
"log-archive-env", False)
self.backupset.log_redaction = self.input.param("log-redaction", False)
self.backupset.redaction_salt = self.input.param(
"redaction-salt", None)
self.backupset.no_log_output_flag = self.input.param(
"no-log-output-flag", False)
self.backupset.ex_logs_path = self.input.param("ex-logs-path", None)
self.backupset.overwrite_user_env = self.input.param(
"overwrite-user-env", False)
self.backupset.overwrite_passwd_env = self.input.param(
"overwrite-passwd-env", False)
self.backupset.disable_conf_res_restriction = self.input.param(
"disable-conf-res-restriction", None)
self.backupset.force_updates = self.input.param("force-updates", False)
self.backupset.resume = self.input.param("resume", False)
self.backupset.purge = self.input.param("purge", False)
self.backupset.start = self.input.param("start", 1)
self.backupset.end = self.input.param("stop", 1)
self.backupset.number_of_backups = self.input.param(
"number_of_backups", 1)
self.replace_ttl_with = self.input.param("replace-ttl-with", None)
self.backupset.backup_host = self.servers[-1]
self.backupset.name = self.input.param("name", "backup")
self.backupset.filter_keys = self.input.param("filter-keys", "")
self.backupset.random_keys = self.input.param("random_keys", False)
self.backupset.filter_values = self.input.param("filter-values", "")
self.backupset.no_ssl_verify = self.input.param("no-ssl-verify", False)
self.backupset.secure_conn = self.input.param("secure-conn", False)
self.backupset.bk_no_cert = self.input.param("bk-no-cert", False)
self.backupset.rt_no_cert = self.input.param("rt-no-cert", False)
self.backupset.backup_list_name = self.input.param("list-names", None)
self.backupset.backup_incr_backup = self.input.param(
"incr-backup", None)
self.backupset.bucket_backup = self.input.param("bucket-backup", None)
shell = RemoteMachineShellConnection(self.servers[0])
info = shell.extract_remote_info().type.lower()
self.root_path = LINUX_ROOT_PATH
self.wget = "wget"
self.os_name = "linux"
self.tmp_path = "/tmp/"
self.long_help_flag = "--help"
self.short_help_flag = "-h"
self.cygwin_bin_path = ""
self.enable_firewal = False
self.rfc3339_date = "date +%s --date='{0} seconds' | ".format(self.replace_ttl_with) + \
"xargs -I {} date --date='@{}' --rfc-3339=seconds | "\
"sed 's/ /T/'"
self.seconds_with_ttl = "date +%s --date='{0} seconds'".format(
self.replace_ttl_with)
if info == 'linux':
if self.nonroot:
base_path = "/home/{0}".format(self.master.ssh_username)
self.database_path = "{0}{1}".format(base_path,
COUCHBASE_DATA_PATH)
self.root_path = "/home/{0}/".format(self.master.ssh_username)
elif info == 'windows':
self.os_name = "windows"
self.cmd_ext = ".exe"
self.wget = "/cygdrive/c/automation/wget.exe"
self.database_path = WIN_COUCHBASE_DATA_PATH_RAW
self.root_path = WIN_ROOT_PATH
self.tmp_path = WIN_TMP_PATH
self.long_help_flag = "help"
self.short_help_flag = "h"
self.cygwin_bin_path = WIN_CYGWIN_BIN_PATH
self.rfc3339_date = "date +%s --date='{0} seconds' | ".format(self.replace_ttl_with) + \
"{0}xargs -I {{}} date --date=\"@'{{}}'\" --rfc-3339=seconds | "\
.format(self.cygwin_bin_path) + \
"sed 's/ /T/'"
win_format = "C:/Program Files"
cygwin_format = "/cygdrive/c/Program\ Files"
if win_format in self.cli_command_location:
self.cli_command_location = self.cli_command_location.replace(
win_format, cygwin_format)
self.backupset.directory = self.input.param(
"dir", WIN_TMP_PATH_RAW + "entbackup")
elif info == 'mac':
self.backupset.directory = self.input.param(
"dir", "/tmp/entbackup")
else:
raise Exception("OS not supported.")
self.non_master_host = self.input.param("non-master", False)
self.value_size = self.input.param("value_size", 512)
self.no_progress_bar = self.input.param("no-progress-bar", True)
self.multi_threads = self.input.param("multi_threads", False)
self.threads_count = self.input.param("threads_count", 1)
self.bucket_delete = self.input.param("bucket_delete", False)
self.bucket_flush = self.input.param("bucket_flush", False)
self.commit = self.input.param("commit", True)
self.ops_type = self.input.param("ops_type", "create")
self.num_threads = self.input.param("num_threads", 5)
self.bk_with_ttl = self.input.param("bk-with-ttl", None)
self.create_fts_index = self.input.param("create-fts-index", False)
self.reset_restore_cluster = self.input.param("reset-restore-cluster",
False)
self.backupset.user_env_with_prompt = \
self.input.param("user-env-with-prompt", False)
self.backupset.passwd_env_with_prompt = \
self.input.param("passwd-env-with-prompt", False)
self.restore_compression_mode = self.input.param(
"restore-compression-mode", None)
self.force_version_upgrade = self.input.param("force-version-upgrade",
None)
self.skip_buckets = self.input.param("skip_buckets", False)
self.num_replicas = self.input.param("replicas", 2)
self.restore_only = self.input.param("restore-only", False)
if self.non_master_host:
self.backupset.cluster_host = self.servers[1]
self.backupset.cluster_host_username = self.servers[
1].rest_username
self.backupset.cluster_host_password = self.servers[
1].rest_password
else:
self.backupset.cluster_host = self.servers[0]
self.backupset.cluster_host_username = self.servers[
0].rest_username
self.backupset.cluster_host_password = self.servers[
0].rest_password
self.same_cluster = self.input.param("same-cluster", False)
if self.same_cluster:
self.backupset.restore_cluster_host = self.input.clusters[0][0]
self.backupset.restore_cluster_host_username = self.input.clusters[
0][0].rest_username
self.backupset.restore_cluster_host_password = self.input.clusters[
0][0].rest_password
else:
self.backupset.restore_cluster_host = self.input.clusters[1][0]
self.backupset.restore_cluster_host_username = self.input.clusters[
1][0].rest_username
self.backupset.restore_cluster_host_password = self.input.clusters[
1][0].rest_password
""" new user to test RBAC """
self.cluster_new_user = self.input.param("new_user", None)
if self.cluster_new_user:
self.backupset.cluster_host_username = self.cluster_new_user
self.backupset.restore_cluster_host_username = self.cluster_new_user
self.backups = []
self.number_of_backups_taken = 0
for cluster in self.clusters:
cluster_util = ClusterUtils(cluster, self.taskmgr)
cluster_util.add_all_nodes_then_rebalance(cluster.servers[1:])
def setUp(self, add_default_cbas_node=True):
super(CBASBaseTest, self).setUp()
if self._testMethodDoc:
self.log.info("Starting Test: %s - %s" %
(self._testMethodName, self._testMethodDoc))
else:
self.log.info("Starting Test: %s" % self._testMethodName)
invalid_ip = '10.111.151.109'
self.cb_bucket_name = self.input.param('cb_bucket_name',
'travel-sample')
self.cbas_bucket_name = self.input.param('cbas_bucket_name', 'travel')
self.cb_bucket_password = self.input.param('cb_bucket_password', None)
self.cb_server_ip = self.input.param("cb_server_ip", None)
self.cb_server_ip = \
self.cb_server_ip.replace('INVALID_IP', invalid_ip) \
if self.cb_server_ip is not None else None
self.cbas_dataset_name = self.input.param("cbas_dataset_name",
'travel_ds')
self.cbas_bucket_name_invalid = \
self.input.param('cbas_bucket_name_invalid', self.cbas_bucket_name)
self.cbas_dataset2_name = self.input.param('cbas_dataset2_name', None)
self.skip_create_dataset = self.input.param('skip_create_dataset',
False)
self.disconnect_if_connected = \
self.input.param('disconnect_if_connected', False)
self.cbas_dataset_name_invalid = \
self.input.param('cbas_dataset_name_invalid',
self.cbas_dataset_name)
self.skip_drop_connection = self.input.param('skip_drop_connection',
False)
self.skip_drop_dataset = self.input.param('skip_drop_dataset', False)
self.query_id = self.input.param('query_id', None)
self.mode = self.input.param('mode', None)
self.num_concurrent_queries = self.input.param('num_queries', 5000)
self.concurrent_batch_size = self.input.param('concurrent_batch_size',
100)
self.compiler_param = self.input.param('compiler_param', None)
self.compiler_param_val = self.input.param('compiler_param_val', None)
self.expect_reject = self.input.param('expect_reject', False)
self.expect_failure = self.input.param('expect_failure', False)
self.compress_dataset = self.input.param('compress_dataset', False)
self.index_name = self.input.param('index_name', "NoName")
self.index_fields = self.input.param('index_fields', None)
if self.index_fields:
self.index_fields = self.index_fields.split("-")
self.retry_time = self.input.param("retry_time", 300)
self.num_retries = self.input.param("num_retries", 1)
self.sample_bucket_dict = {
TravelSample().name: TravelSample(),
BeerSample().name: BeerSample()
}
self.sample_bucket = None
self.flush_enabled = Bucket.FlushBucket.ENABLED
self.test_abort_snapshot = self.input.param("test_abort_snapshot",
False)
if hasattr(self, "cluster"):
self._cb_cluster = self.cluster
else:
self._cb_cluster = self.get_clusters()
self.expected_error = self.input.param("error", None)
self.spec_name = self.input.param("bucket_spec", None)
# Single cluster support
if hasattr(self, "cluster"):
for server in self.servers:
if "cbas" in server.services:
self.cluster.cbas_nodes.append(server)
if "kv" in server.services:
self.cluster.kv_nodes.append(server)
rest = RestConnection(server)
rest.set_data_path(data_path=server.data_path,
index_path=server.index_path,
cbas_path=server.cbas_path)
if self.expected_error:
self.expected_error = \
self.expected_error.replace("INVALID_IP", invalid_ip)
self.expected_error = \
self.expected_error.replace("PORT",
self.cluster.master.port)
self.otpNodes = []
self.cbas_path = server.cbas_path
self.rest = RestConnection(self.cluster.master)
self.log.info(
"Setting the min possible memory quota so that adding "
"more nodes to the cluster wouldn't be a problem.")
self.rest.set_service_memoryQuota(service='memoryQuota',
memoryQuota=MIN_KV_QUOTA)
self.rest.set_service_memoryQuota(service='ftsMemoryQuota',
memoryQuota=FTS_QUOTA)
self.rest.set_service_memoryQuota(service='indexMemoryQuota',
memoryQuota=INDEX_QUOTA)
self.set_cbas_memory_from_available_free_memory = \
self.input.param('set_cbas_memory_from_available_free_memory',
False)
if self.expected_error:
self.expected_error = \
self.expected_error.replace("INVALID_IP", invalid_ip)
self.expected_error = \
self.expected_error.replace("PORT",
self.cluster.master.port)
if self.set_cbas_memory_from_available_free_memory:
info = self.rest.get_nodes_self()
self.cbas_memory_quota = int(
(info.memoryFree // 1024**2) * 0.9)
self.log.info("Setting %d memory quota for CBAS" %
self.cbas_memory_quota)
self.rest.set_service_memoryQuota(
service='cbasMemoryQuota',
memoryQuota=self.cbas_memory_quota)
else:
self.log.info("Setting %d memory quota for CBAS" % CBAS_QUOTA)
self.cbas_memory_quota = CBAS_QUOTA
self.rest.set_service_memoryQuota(service='cbasMemoryQuota',
memoryQuota=CBAS_QUOTA)
self.cbas_util = None
if self.cluster.cbas_nodes:
self.cbas_node = self.cluster.cbas_nodes[0]
self.cbas_util = CbasUtil(self.cluster.master, self.cbas_node)
if "cbas" in self.cluster.master.services:
self.cleanup_cbas()
if add_default_cbas_node:
if self.cluster.master.ip != self.cbas_node.ip:
self.otpNodes.append(
ClusterUtils(self.cluster,
self.task_manager).add_node(
self.cbas_node))
else:
self.otpNodes = self.rest.node_statuses()
''' This cbas cleanup is actually not needed.
When a node is added to the cluster,
it is automatically cleaned-up.'''
self.cleanup_cbas()
self.cluster.cbas_nodes.remove(self.cbas_node)
if self.spec_name is not None:
try:
self.collectionSetUp(self.cluster, self.bucket_util,
self.cluster_util)
except Java_base_exception as exception:
self.handle_collection_setup_exception(exception)
except Exception as exception:
self.handle_collection_setup_exception(exception)
else:
if self.default_bucket:
self.bucket_util.create_default_bucket(
bucket_type=self.bucket_type,
ram_quota=self.bucket_size,
replica=self.num_replicas,
conflict_resolution=self.
bucket_conflict_resolution_type,
replica_index=self.bucket_replica_index,
storage=self.bucket_storage,
eviction_policy=self.bucket_eviction_policy,
flush_enabled=self.flush_enabled)
elif self.cb_bucket_name in self.sample_bucket_dict.keys():
self.sample_bucket = \
self.sample_bucket_dict[self.cb_bucket_name]
else:
# Multi Cluster Support
for cluster in self._cb_cluster:
cluster.cluster_util = ClusterUtils(cluster, self.task_manager)
cluster.bucket_util = BucketUtils(cluster,
cluster.cluster_util,
self.task)
for server in cluster.servers:
if "cbas" in server.services:
cluster.cbas_nodes.append(server)
if "kv" in server.services:
cluster.kv_nodes.append(server)
rest = RestConnection(server)
rest.set_data_path(data_path=server.data_path,
index_path=server.index_path,
cbas_path=server.cbas_path)
if self.expected_error:
cluster.expected_error = \
self.expected_error.replace("INVALID_IP", invalid_ip)
cluster.expected_error = \
self.expected_error.replace("PORT",
cluster.master.port)
cluster.otpNodes = list()
cluster.cbas_path = server.cbas_path
cluster.rest = RestConnection(cluster.master)
self.log.info(
"Setting the min possible memory quota so that adding "
"more nodes to the cluster wouldn't be a problem.")
cluster.rest.set_service_memoryQuota(service='memoryQuota',
memoryQuota=MIN_KV_QUOTA)
cluster.rest.set_service_memoryQuota(service='ftsMemoryQuota',
memoryQuota=FTS_QUOTA)
cluster.rest.set_service_memoryQuota(
service='indexMemoryQuota', memoryQuota=INDEX_QUOTA)
cluster.set_cbas_memory_from_available_free_memory = \
self.input.param(
'set_cbas_memory_from_available_free_memory', False)
if cluster.set_cbas_memory_from_available_free_memory:
info = cluster.rest.get_nodes_self()
cluster.cbas_memory_quota = int(
(info.memoryFree // 1024**2) * 0.9)
self.log.info("Setting %d memory quota for CBAS" %
cluster.cbas_memory_quota)
cluster.rest.set_service_memoryQuota(
service='cbasMemoryQuota',
memoryQuota=cluster.cbas_memory_quota)
else:
self.log.info("Setting %d memory quota for CBAS" %
CBAS_QUOTA)
cluster.cbas_memory_quota = CBAS_QUOTA
cluster.rest.set_service_memoryQuota(
service='cbasMemoryQuota', memoryQuota=CBAS_QUOTA)
cluster.cbas_util = None
# Drop any existing buckets and datasets
if cluster.cbas_nodes:
cluster.cbas_node = cluster.cbas_nodes[0]
cluster.cbas_util = CbasUtil(cluster.master,
cluster.cbas_node, self.task)
if "cbas" in cluster.master.services:
self.cleanup_cbas(cluster.cbas_util)
if add_default_cbas_node:
if cluster.master.ip != cluster.cbas_node.ip:
cluster.otpNodes.append(
cluster.cluster_util.add_node(
cluster.cbas_node))
else:
cluster.otpNodes = cluster.rest.node_statuses()
"""
This cbas cleanup is actually not needed.
When a node is added to the cluster,
it is automatically cleaned-up.
"""
self.cleanup_cbas(cluster.cbas_util)
cluster.cbas_nodes.remove(cluster.cbas_node)
if self.spec_name is not None:
try:
self.collectionSetUp(cluster, cluster.bucket_util,
cluster.cluster_util)
except Java_base_exception as exception:
self.handle_collection_setup_exception(exception)
except Exception as exception:
self.handle_collection_setup_exception(exception)
else:
if self.default_bucket:
cluster.bucket_util.create_default_bucket(
bucket_type=self.bucket_type,
ram_quota=self.bucket_size,
replica=self.num_replicas,
conflict_resolution=self.
bucket_conflict_resolution_type,
replica_index=self.bucket_replica_index,
storage=self.bucket_storage,
eviction_policy=self.bucket_eviction_policy,
flush_enabled=self.flush_enabled)
elif self.cb_bucket_name in self.sample_bucket_dict.keys():
self.sample_bucket = self.sample_bucket_dict[
self.cb_bucket_name]
cluster.bucket_util.add_rbac_user()
self.log.info(
"=== CBAS_BASE setup was finished for test #{0} {1} ===".format(
self.case_number, self._testMethodName))
def setUp(self, add_default_cbas_node=True):
super(CBASBaseTest, self).setUp()
if self._testMethodDoc:
self.log.info("Starting Test: %s - %s" %
(self._testMethodName, self._testMethodDoc))
else:
self.log.info("Starting Test: %s" % self._testMethodName)
invalid_ip = '10.111.151.109'
self.cb_bucket_name = self.input.param('cb_bucket_name',
'travel-sample')
self.cbas_bucket_name = self.input.param('cbas_bucket_name', 'travel')
self.cb_bucket_password = self.input.param('cb_bucket_password', None)
self.cb_server_ip = self.input.param("cb_server_ip", None)
self.cb_server_ip = \
self.cb_server_ip.replace('INVALID_IP', invalid_ip) \
if self.cb_server_ip is not None else None
self.cbas_dataset_name = self.input.param("cbas_dataset_name",
'travel_ds')
self.cbas_bucket_name_invalid = \
self.input.param('cbas_bucket_name_invalid', self.cbas_bucket_name)
self.cbas_dataset2_name = self.input.param('cbas_dataset2_name', None)
self.skip_create_dataset = self.input.param('skip_create_dataset',
False)
self.disconnect_if_connected = \
self.input.param('disconnect_if_connected', False)
self.cbas_dataset_name_invalid = \
self.input.param('cbas_dataset_name_invalid',
self.cbas_dataset_name)
self.skip_drop_connection = self.input.param('skip_drop_connection',
False)
self.skip_drop_dataset = self.input.param('skip_drop_dataset', False)
self.query_id = self.input.param('query_id', None)
self.mode = self.input.param('mode', None)
self.num_concurrent_queries = self.input.param('num_queries', 5000)
self.concurrent_batch_size = self.input.param('concurrent_batch_size',
100)
self.compiler_param = self.input.param('compiler_param', None)
self.compiler_param_val = self.input.param('compiler_param_val', None)
self.expect_reject = self.input.param('expect_reject', False)
self.expect_failure = self.input.param('expect_failure', False)
self.compress_dataset = self.input.param('compress_dataset', False)
self.index_name = self.input.param('index_name', "NoName")
self.index_fields = self.input.param('index_fields', None)
if self.index_fields:
self.index_fields = self.index_fields.split("-")
self.retry_time = self.input.param("retry_time", 300)
self.num_retries = self.input.param("num_retries", 1)
self.sample_bucket_dict = {
TravelSample().name: TravelSample(),
BeerSample().name: BeerSample()
}
self.sample_bucket = None
self.flush_enabled = Bucket.FlushBucket.ENABLED
self.test_abort_snapshot = self.input.param("test_abort_snapshot",
False)
self.cbas_spec_name = self.input.param("cbas_spec", None)
self._cb_cluster = self.get_clusters()
self.expected_error = self.input.param("error", None)
self.bucket_spec = self.input.param("bucket_spec", None)
self.doc_spec_name = self.input.param("doc_spec_name", "initial_load")
self.set_cbas_memory_from_available_free_memory = self.input.param(
'set_cbas_memory_from_available_free_memory', False)
self.parallel_load_percent = int(
self.input.param("parallel_load_percent", 0))
self.cbas_kill_count = self.input.param("cbas_kill_count", 0)
self.memcached_kill_count = self.input.param("memcached_kill_count", 0)
self.tamper_links_count = self.input.param("tamper_links_count", 0)
self.cbas_node = None
services = None
nodes_init = None
# Single cluster support
if len(self._cb_cluster) == 1:
self._cb_cluster = self._cb_cluster[0]
self.cluster.nodes_in_cluster.extend([self.cluster.master])
if self.services_init and self.nodes_init >= 3:
if len(self.cluster.servers) < self.nodes_init or \
len(self.services_init.split("-")) != self.nodes_init:
self.fail("Configuration error. Re-check nodes_init, "
"services_init in .conf file and servers "
"available in .ini "
"file")
services = list()
for service in self.services_init.split(
"-")[1:self.nodes_init]:
services.append(service.replace(":", ","))
# Initialize cluster using given nodes
nodes_init = list(
filter(lambda node: node.ip != self.cluster.master.ip,
self.cluster.servers[1:self.nodes_init]))
for node, services_init in map(None, nodes_init, services):
if services_init is None:
services.append("kv")
if not self.cbas_node and "cbas" in services_init:
self.cbas_node = node
self.cbas_node.services = services_init
idx = self.cluster.servers.index(node)
self.cluster.servers[idx].services = services_init
for server in self.cluster.servers:
if "cbas" in server.services:
self.cluster.cbas_nodes.append(server)
if "kv" in server.services:
self.cluster.kv_nodes.append(server)
rest = RestConnection(server)
rest.set_data_path(data_path=server.data_path,
index_path=server.index_path,
cbas_path=server.cbas_path)
if self.expected_error:
self.expected_error = \
self.expected_error.replace("INVALID_IP", invalid_ip)
self.expected_error = \
self.expected_error.replace("PORT",
self.cluster.master.port)
self.otpNodes = []
self.cbas_path = server.cbas_path
self.rest = RestConnection(self.cluster.master)
if not self.set_cbas_memory_from_available_free_memory:
self.log.info(
"Setting the min possible memory quota so that adding "
"more nodes to the cluster wouldn't be a problem.")
self.rest.set_service_memoryQuota(service='memoryQuota',
memoryQuota=MIN_KV_QUOTA)
self.rest.set_service_memoryQuota(service='ftsMemoryQuota',
memoryQuota=FTS_QUOTA)
self.rest.set_service_memoryQuota(service='indexMemoryQuota',
memoryQuota=INDEX_QUOTA)
self.set_cbas_memory_from_available_free_memory = \
self.input.param(
'set_cbas_memory_from_available_free_memory', False)
self.log.info("Setting %d memory quota for CBAS" % CBAS_QUOTA)
self.cbas_memory_quota = CBAS_QUOTA
self.rest.set_service_memoryQuota(service='cbasMemoryQuota',
memoryQuota=CBAS_QUOTA)
if self.expected_error:
self.expected_error = \
self.expected_error.replace("INVALID_IP", invalid_ip)
self.expected_error = \
self.expected_error.replace("PORT",
self.cluster.master.port)
self.cbas_util = None
if self.cluster.cbas_nodes:
if not self.cbas_node:
available_cbas_nodes = list(
filter(lambda node: node.ip != self.cluster.master.ip,
self.cluster.cbas_nodes))
self.cbas_node = available_cbas_nodes[0]
if self.set_cbas_memory_from_available_free_memory:
self.set_memory_for_services(self.rest, self.cluster_util,
self.cbas_node,
self.cbas_node.services)
self.cbas_util = CbasUtil(self.cluster.master, self.cbas_node)
self.cbas_util_v2 = CbasUtilV2(self.cluster.master,
self.cbas_node, self.task)
if "cbas" in self.cluster.master.services:
self.cleanup_cbas()
if add_default_cbas_node:
if self.cluster.master.ip != self.cbas_node.ip:
self.otpNodes.append(
ClusterUtils(self.cluster,
self.task_manager).add_node(
self.cbas_node))
self.cluster.nodes_in_cluster.append(self.cbas_node)
if nodes_init:
idx = nodes_init.index(self.cbas_node)
services.pop(idx)
nodes_init.remove(self.cbas_node)
else:
self.otpNodes = self.rest.node_statuses()
''' This cbas cleanup is actually not needed.
When a node is added to the cluster,
it is automatically cleaned-up.'''
self.cleanup_cbas()
self.cluster.cbas_nodes.remove(self.cbas_node)
if nodes_init:
self.task.rebalance([self.cluster.master],
nodes_init, [],
services=services)
self.cluster.nodes_in_cluster.extend(nodes_init)
if self.bucket_spec is not None:
try:
self.collectionSetUp(self.cluster, self.bucket_util,
self.cluster_util)
except Java_base_exception as exception:
self.handle_collection_setup_exception(exception)
except Exception as exception:
self.handle_collection_setup_exception(exception)
else:
if self.default_bucket:
self.bucket_util.create_default_bucket(
bucket_type=self.bucket_type,
ram_quota=self.bucket_size,
replica=self.num_replicas,
conflict_resolution=self.
bucket_conflict_resolution_type,
replica_index=self.bucket_replica_index,
storage=self.bucket_storage,
eviction_policy=self.bucket_eviction_policy,
flush_enabled=self.flush_enabled)
elif self.cb_bucket_name in self.sample_bucket_dict.keys():
self.sample_bucket = \
self.sample_bucket_dict[self.cb_bucket_name]
elif len(self._cb_cluster) > 1:
# Multi Cluster Support
for cluster in self._cb_cluster:
cluster.cluster_util = ClusterUtils(cluster, self.task_manager)
cluster.bucket_util = BucketUtils(cluster,
cluster.cluster_util,
self.task)
for server in cluster.servers:
if "cbas" in server.services:
cluster.cbas_nodes.append(server)
if "kv" in server.services:
cluster.kv_nodes.append(server)
rest = RestConnection(server)
rest.set_data_path(data_path=server.data_path,
index_path=server.index_path,
cbas_path=server.cbas_path)
if self.expected_error:
cluster.expected_error = \
self.expected_error.replace("INVALID_IP", invalid_ip)
cluster.expected_error = \
self.expected_error.replace("PORT",
cluster.master.port)
cluster.otpNodes = list()
cluster.cbas_path = server.cbas_path
cluster.rest = RestConnection(cluster.master)
if not self.set_cbas_memory_from_available_free_memory:
self.log.info(
"Setting the min possible memory quota so that adding "
"more nodes to the cluster wouldn't be a problem.")
cluster.rest.set_service_memoryQuota(
service='memoryQuota', memoryQuota=MIN_KV_QUOTA)
cluster.rest.set_service_memoryQuota(
service='ftsMemoryQuota', memoryQuota=FTS_QUOTA)
cluster.rest.set_service_memoryQuota(
service='indexMemoryQuota', memoryQuota=INDEX_QUOTA)
cluster.set_cbas_memory_from_available_free_memory = \
self.input.param(
'set_cbas_memory_from_available_free_memory', False)
self.log.info("Setting %d memory quota for CBAS" %
CBAS_QUOTA)
cluster.cbas_memory_quota = CBAS_QUOTA
cluster.rest.set_service_memoryQuota(
service='cbasMemoryQuota', memoryQuota=CBAS_QUOTA)
cluster.cbas_util = None
# Drop any existing buckets and datasets
if cluster.cbas_nodes:
cluster.cbas_node = cluster.cbas_nodes[0]
if self.set_cbas_memory_from_available_free_memory:
self.set_memory_for_services(
cluster.rest, cluster.cluster_util,
cluster.cbas_node, cluster.cbas_node.services)
cluster.cbas_util = CbasUtil(cluster.master,
cluster.cbas_node, self.task)
cluster.cbas_util_v2 = CbasUtilV2(cluster.master,
cluster.cbas_node)
if "cbas" in cluster.master.services:
self.cleanup_cbas(cluster.cbas_util)
if add_default_cbas_node:
if cluster.master.ip != cluster.cbas_node.ip:
cluster.otpNodes.append(
cluster.cluster_util.add_node(
cluster.cbas_node))
else:
cluster.otpNodes = cluster.rest.node_statuses()
"""
This cbas cleanup is actually not needed.
When a node is added to the cluster,
it is automatically cleaned-up.
"""
self.cleanup_cbas(cluster.cbas_util)
cluster.cbas_nodes.remove(cluster.cbas_node)
if self.bucket_spec is not None:
try:
self.collectionSetUp(cluster, cluster.bucket_util,
cluster.cluster_util)
except Java_base_exception as exception:
self.handle_collection_setup_exception(exception)
except Exception as exception:
self.handle_collection_setup_exception(exception)
else:
if self.default_bucket:
cluster.bucket_util.create_default_bucket(
bucket_type=self.bucket_type,
ram_quota=self.bucket_size,
replica=self.num_replicas,
conflict_resolution=self.
bucket_conflict_resolution_type,
replica_index=self.bucket_replica_index,
storage=self.bucket_storage,
eviction_policy=self.bucket_eviction_policy,
flush_enabled=self.flush_enabled)
elif self.cb_bucket_name in self.sample_bucket_dict.keys():
self.sample_bucket = self.sample_bucket_dict[
self.cb_bucket_name]
cluster.bucket_util.add_rbac_user()
else:
self.fail("No cluster is available")
self.log.info(
"=== CBAS_BASE setup was finished for test #{0} {1} ===".format(
self.case_number, self._testMethodName))
def setUp(self):
super(OnPremBaseTest, self).setUp()
# Framework specific parameters (Extension from cb_basetest)
self.skip_cluster_reset = self.input.param("skip_cluster_reset", False)
self.skip_setup_cleanup = self.input.param("skip_setup_cleanup", False)
# End of framework parameters
# Cluster level info settings
self.log_info = self.input.param("log_info", None)
self.log_location = self.input.param("log_location", None)
self.stat_info = self.input.param("stat_info", None)
self.port = self.input.param("port", None)
self.port_info = self.input.param("port_info", None)
self.servers = self.input.servers
self.num_servers = self.input.param("servers", len(self.servers))
self.vbuckets = self.input.param("vbuckets", CbServer.total_vbuckets)
self.gsi_type = self.input.param("gsi_type", 'plasma')
# Memory quota settings
# Max memory quota to utilize per node
self.quota_percent = self.input.param("quota_percent", 100)
# Services' RAM quota to set on cluster
self.kv_mem_quota_percent = self.input.param("kv_quota_percent", None)
self.index_mem_quota_percent = \
self.input.param("index_quota_percent", None)
self.fts_mem_quota_percent = \
self.input.param("fts_quota_percent", None)
self.cbas_mem_quota_percent = \
self.input.param("cbas_quota_percent", None)
self.eventing_mem_quota_percent = \
self.input.param("eventing_quota_percent", None)
# CBAS setting
self.jre_path = self.input.param("jre_path", None)
self.enable_dp = self.input.param("enable_dp", False)
# End of cluster info parameters
# Bucket specific params
# Note: Over riding bucket_eviction_policy from CouchbaseBaseTest
self.bucket_eviction_policy = \
self.input.param("bucket_eviction_policy",
Bucket.EvictionPolicy.VALUE_ONLY)
self.bucket_replica_index = self.input.param("bucket_replica_index", 1)
if self.bucket_storage == Bucket.StorageBackend.magma:
self.bucket_eviction_policy = Bucket.EvictionPolicy.FULL_EVICTION
# End of bucket parameters
self.services_in = self.input.param("services_in", None)
self.forceEject = self.input.param("forceEject", False)
self.wait_timeout = self.input.param("wait_timeout", 120)
self.verify_unacked_bytes = \
self.input.param("verify_unacked_bytes", False)
self.disabled_consistent_view = \
self.input.param("disabled_consistent_view", None)
self.rebalanceIndexWaitingDisabled = \
self.input.param("rebalanceIndexWaitingDisabled", None)
self.rebalanceIndexPausingDisabled = \
self.input.param("rebalanceIndexPausingDisabled", None)
self.maxParallelIndexers = \
self.input.param("maxParallelIndexers", None)
self.maxParallelReplicaIndexers = \
self.input.param("maxParallelReplicaIndexers", None)
self.use_https = self.input.param("use_https", False)
self.enforce_tls = self.input.param("enforce_tls", False)
self.ipv4_only = self.input.param("ipv4_only", False)
self.ipv6_only = self.input.param("ipv6_only", False)
self.multiple_ca = self.input.param("multiple_ca", False)
if self.use_https:
CbServer.use_https = True
trust_all_certs()
self.node_utils.cleanup_pcaps(self.servers)
self.collect_pcaps = self.input.param("collect_pcaps", False)
if self.collect_pcaps:
self.node_utils.start_collect_pcaps(self.servers)
'''
Be careful while using this flag.
This is only and only for stand-alone tests.
During bugs reproductions, when a crash is seen
stop_server_on_crash will stop the server
so that we can collect data/logs/dumps at the right time
'''
self.stop_server_on_crash = self.input.param("stop_server_on_crash",
False)
self.collect_data = self.input.param("collect_data", False)
self.validate_system_event_logs = \
self.input.param("validate_sys_event_logs", False)
self.nonroot = False
self.crash_warning = self.input.param("crash_warning", False)
# Populate memcached_port in case of cluster_run
cluster_run_base_port = ClusterRun.port
if int(self.input.servers[0].port) == ClusterRun.port:
for server in self.input.servers:
server.port = cluster_run_base_port
cluster_run_base_port += 1
# If not defined in node.ini under 'memcached_port' section
if server.memcached_port is CbServer.memcached_port:
server.memcached_port = \
ClusterRun.memcached_port \
+ (2 * (int(server.port) - ClusterRun.port))
self.log_setup_status(self.__class__.__name__, "started")
cluster_name_format = "C%s"
default_cluster_index = counter_index = 1
if len(self.input.clusters) > 1:
# Multi cluster setup
for _, nodes in self.input.clusters.iteritems():
cluster_name = cluster_name_format % counter_index
tem_cluster = CBCluster(name=cluster_name,
servers=nodes,
vbuckets=self.vbuckets)
self.cb_clusters[cluster_name] = tem_cluster
counter_index += 1
else:
# Single cluster
cluster_name = cluster_name_format % counter_index
self.cb_clusters[cluster_name] = CBCluster(name=cluster_name,
servers=self.servers,
vbuckets=self.vbuckets)
# Initialize self.cluster with first available cluster as default
self.cluster = self.cb_clusters[cluster_name_format %
default_cluster_index]
self.cluster_util = ClusterUtils(self.task_manager)
self.bucket_util = BucketUtils(self.cluster_util, self.task)
CbServer.enterprise_edition = \
self.cluster_util.is_enterprise_edition(self.cluster)
if CbServer.enterprise_edition:
self.cluster.edition = "enterprise"
else:
self.cluster.edition = "community"
if self.standard_buckets > 10:
self.bucket_util.change_max_buckets(self.cluster.master,
self.standard_buckets)
for cluster_name, cluster in self.cb_clusters.items():
# Append initial master node to the nodes_in_cluster list
cluster.nodes_in_cluster.append(cluster.master)
shell = RemoteMachineShellConnection(cluster.master)
self.os_info = shell.extract_remote_info().type.lower()
if self.os_info != 'windows':
if cluster.master.ssh_username != "root":
self.nonroot = True
shell.disconnect()
break
shell.disconnect()
self.log_setup_status("OnPremBaseTest", "started")
try:
# Construct dict of mem. quota percent / mb per service
mem_quota_percent = dict()
# Construct dict of mem. quota percent per service
if self.kv_mem_quota_percent:
mem_quota_percent[CbServer.Services.KV] = \
self.kv_mem_quota_percent
if self.index_mem_quota_percent:
mem_quota_percent[CbServer.Services.INDEX] = \
self.index_mem_quota_percent
if self.cbas_mem_quota_percent:
mem_quota_percent[CbServer.Services.CBAS] = \
self.cbas_mem_quota_percent
if self.fts_mem_quota_percent:
mem_quota_percent[CbServer.Services.FTS] = \
self.fts_mem_quota_percent
if self.eventing_mem_quota_percent:
mem_quota_percent[CbServer.Services.EVENTING] = \
self.eventing_mem_quota_percent
if not mem_quota_percent:
mem_quota_percent = None
if self.skip_setup_cleanup:
# Update current server/service map and buckets for the cluster
for _, cluster in self.cb_clusters.items():
self.cluster_util.update_cluster_nodes_service_list(
cluster)
cluster.buckets = self.bucket_util.get_all_buckets(cluster)
return
else:
for cluster_name, cluster in self.cb_clusters.items():
self.log.info("Delete all buckets and rebalance out "
"other nodes from '%s'" % cluster_name)
self.cluster_util.cluster_cleanup(cluster,
self.bucket_util)
reload(Cb_constants)
# avoid clean up if the previous test has been tear down
if self.case_number == 1 or self.case_number > 1000:
if self.case_number > 1000:
self.log.warn("TearDown for prev test failed. Will retry")
self.case_number -= 1000
self.tearDownEverything(reset_cluster_env_vars=False)
for cluster_name, cluster in self.cb_clusters.items():
if not self.skip_cluster_reset:
self.initialize_cluster(
cluster_name,
cluster,
services=None,
services_mem_quota_percent=mem_quota_percent)
# Update initial service map for the master node
self.cluster_util.update_cluster_nodes_service_list(cluster)
# Set this unconditionally
RestConnection(cluster.master).set_internalSetting(
"magmaMinMemoryQuota", 256)
# Enable dp_version since we need collections enabled
if self.enable_dp:
tasks = []
for server in self.cluster.servers:
task = self.node_utils.async_enable_dp(server)
tasks.append(task)
for task in tasks:
self.task_manager.get_task_result(task)
# Enforce tls on nodes of all clusters
if self.use_https and self.enforce_tls:
for _, cluster in self.cb_clusters.items():
tasks = []
for node in cluster.servers:
task = self.node_utils.async_enable_tls(node)
tasks.append(task)
for task in tasks:
self.task_manager.get_task_result(task)
self.log.info(
"Validating if services obey tls only on servers {0}".
format(cluster.servers))
status = self.cluster_util.check_if_services_obey_tls(
cluster.servers)
if not status:
self.fail("Services did not honor enforce tls")
# Enforce IPv4 or IPv6 or both
if self.ipv4_only or self.ipv6_only:
for _, cluster in self.cb_clusters.items():
status, msg = self.cluster_util.enable_disable_ip_address_family_type(
cluster, True, self.ipv4_only, self.ipv6_only)
if not status:
self.fail(msg)
self.standard = self.input.param("standard", "pkcs8")
self.passphrase_type = self.input.param("passphrase_type",
"script")
self.encryption_type = self.input.param("encryption_type",
"aes256")
if self.multiple_ca:
for _, cluster in self.cb_clusters.items():
cluster.x509 = x509main(
host=cluster.master,
standard=self.standard,
encryption_type=self.encryption_type,
passphrase_type=self.passphrase_type)
self.generate_and_upload_cert(cluster.servers,
cluster.x509,
upload_root_certs=True,
upload_node_certs=True,
upload_client_certs=True)
payload = "name=cbadminbucket&roles=admin&password=password"
rest = RestConnection(cluster.master)
rest.add_set_builtin_user("cbadminbucket", payload)
for cluster_name, cluster in self.cb_clusters.items():
self.modify_cluster_settings(cluster)
# Track test start time only if we need system log validation
if self.validate_system_event_logs:
self.system_events.set_test_start_time()
self.log_setup_status("OnPremBaseTest", "finished")
self.__log("started")
except Exception as e:
traceback.print_exc()
self.task.shutdown(force=True)
self.fail(e)
finally:
# Track test start time only if we need system log validation
if self.validate_system_event_logs:
self.system_events.set_test_start_time()
self.log_setup_status("OnPremBaseTest", "finished")
class OnPremBaseTest(CouchbaseBaseTest):
def setUp(self):
super(OnPremBaseTest, self).setUp()
# Framework specific parameters (Extension from cb_basetest)
self.skip_cluster_reset = self.input.param("skip_cluster_reset", False)
self.skip_setup_cleanup = self.input.param("skip_setup_cleanup", False)
# End of framework parameters
# Cluster level info settings
self.log_info = self.input.param("log_info", None)
self.log_location = self.input.param("log_location", None)
self.stat_info = self.input.param("stat_info", None)
self.port = self.input.param("port", None)
self.port_info = self.input.param("port_info", None)
self.servers = self.input.servers
self.num_servers = self.input.param("servers", len(self.servers))
self.vbuckets = self.input.param("vbuckets", CbServer.total_vbuckets)
self.gsi_type = self.input.param("gsi_type", 'plasma')
# Memory quota settings
# Max memory quota to utilize per node
self.quota_percent = self.input.param("quota_percent", 100)
# Services' RAM quota to set on cluster
self.kv_mem_quota_percent = self.input.param("kv_quota_percent", None)
self.index_mem_quota_percent = \
self.input.param("index_quota_percent", None)
self.fts_mem_quota_percent = \
self.input.param("fts_quota_percent", None)
self.cbas_mem_quota_percent = \
self.input.param("cbas_quota_percent", None)
self.eventing_mem_quota_percent = \
self.input.param("eventing_quota_percent", None)
# CBAS setting
self.jre_path = self.input.param("jre_path", None)
self.enable_dp = self.input.param("enable_dp", False)
# End of cluster info parameters
# Bucket specific params
# Note: Over riding bucket_eviction_policy from CouchbaseBaseTest
self.bucket_eviction_policy = \
self.input.param("bucket_eviction_policy",
Bucket.EvictionPolicy.VALUE_ONLY)
self.bucket_replica_index = self.input.param("bucket_replica_index", 1)
if self.bucket_storage == Bucket.StorageBackend.magma:
self.bucket_eviction_policy = Bucket.EvictionPolicy.FULL_EVICTION
# End of bucket parameters
self.services_in = self.input.param("services_in", None)
self.forceEject = self.input.param("forceEject", False)
self.wait_timeout = self.input.param("wait_timeout", 120)
self.verify_unacked_bytes = \
self.input.param("verify_unacked_bytes", False)
self.disabled_consistent_view = \
self.input.param("disabled_consistent_view", None)
self.rebalanceIndexWaitingDisabled = \
self.input.param("rebalanceIndexWaitingDisabled", None)
self.rebalanceIndexPausingDisabled = \
self.input.param("rebalanceIndexPausingDisabled", None)
self.maxParallelIndexers = \
self.input.param("maxParallelIndexers", None)
self.maxParallelReplicaIndexers = \
self.input.param("maxParallelReplicaIndexers", None)
self.use_https = self.input.param("use_https", False)
self.enforce_tls = self.input.param("enforce_tls", False)
self.ipv4_only = self.input.param("ipv4_only", False)
self.ipv6_only = self.input.param("ipv6_only", False)
self.multiple_ca = self.input.param("multiple_ca", False)
if self.use_https:
CbServer.use_https = True
trust_all_certs()
self.node_utils.cleanup_pcaps(self.servers)
self.collect_pcaps = self.input.param("collect_pcaps", False)
if self.collect_pcaps:
self.node_utils.start_collect_pcaps(self.servers)
'''
Be careful while using this flag.
This is only and only for stand-alone tests.
During bugs reproductions, when a crash is seen
stop_server_on_crash will stop the server
so that we can collect data/logs/dumps at the right time
'''
self.stop_server_on_crash = self.input.param("stop_server_on_crash",
False)
self.collect_data = self.input.param("collect_data", False)
self.validate_system_event_logs = \
self.input.param("validate_sys_event_logs", False)
self.nonroot = False
self.crash_warning = self.input.param("crash_warning", False)
# Populate memcached_port in case of cluster_run
cluster_run_base_port = ClusterRun.port
if int(self.input.servers[0].port) == ClusterRun.port:
for server in self.input.servers:
server.port = cluster_run_base_port
cluster_run_base_port += 1
# If not defined in node.ini under 'memcached_port' section
if server.memcached_port is CbServer.memcached_port:
server.memcached_port = \
ClusterRun.memcached_port \
+ (2 * (int(server.port) - ClusterRun.port))
self.log_setup_status(self.__class__.__name__, "started")
cluster_name_format = "C%s"
default_cluster_index = counter_index = 1
if len(self.input.clusters) > 1:
# Multi cluster setup
for _, nodes in self.input.clusters.iteritems():
cluster_name = cluster_name_format % counter_index
tem_cluster = CBCluster(name=cluster_name,
servers=nodes,
vbuckets=self.vbuckets)
self.cb_clusters[cluster_name] = tem_cluster
counter_index += 1
else:
# Single cluster
cluster_name = cluster_name_format % counter_index
self.cb_clusters[cluster_name] = CBCluster(name=cluster_name,
servers=self.servers,
vbuckets=self.vbuckets)
# Initialize self.cluster with first available cluster as default
self.cluster = self.cb_clusters[cluster_name_format %
default_cluster_index]
self.cluster_util = ClusterUtils(self.task_manager)
self.bucket_util = BucketUtils(self.cluster_util, self.task)
CbServer.enterprise_edition = \
self.cluster_util.is_enterprise_edition(self.cluster)
if CbServer.enterprise_edition:
self.cluster.edition = "enterprise"
else:
self.cluster.edition = "community"
if self.standard_buckets > 10:
self.bucket_util.change_max_buckets(self.cluster.master,
self.standard_buckets)
for cluster_name, cluster in self.cb_clusters.items():
# Append initial master node to the nodes_in_cluster list
cluster.nodes_in_cluster.append(cluster.master)
shell = RemoteMachineShellConnection(cluster.master)
self.os_info = shell.extract_remote_info().type.lower()
if self.os_info != 'windows':
if cluster.master.ssh_username != "root":
self.nonroot = True
shell.disconnect()
break
shell.disconnect()
self.log_setup_status("OnPremBaseTest", "started")
try:
# Construct dict of mem. quota percent / mb per service
mem_quota_percent = dict()
# Construct dict of mem. quota percent per service
if self.kv_mem_quota_percent:
mem_quota_percent[CbServer.Services.KV] = \
self.kv_mem_quota_percent
if self.index_mem_quota_percent:
mem_quota_percent[CbServer.Services.INDEX] = \
self.index_mem_quota_percent
if self.cbas_mem_quota_percent:
mem_quota_percent[CbServer.Services.CBAS] = \
self.cbas_mem_quota_percent
if self.fts_mem_quota_percent:
mem_quota_percent[CbServer.Services.FTS] = \
self.fts_mem_quota_percent
if self.eventing_mem_quota_percent:
mem_quota_percent[CbServer.Services.EVENTING] = \
self.eventing_mem_quota_percent
if not mem_quota_percent:
mem_quota_percent = None
if self.skip_setup_cleanup:
# Update current server/service map and buckets for the cluster
for _, cluster in self.cb_clusters.items():
self.cluster_util.update_cluster_nodes_service_list(
cluster)
cluster.buckets = self.bucket_util.get_all_buckets(cluster)
return
else:
for cluster_name, cluster in self.cb_clusters.items():
self.log.info("Delete all buckets and rebalance out "
"other nodes from '%s'" % cluster_name)
self.cluster_util.cluster_cleanup(cluster,
self.bucket_util)
reload(Cb_constants)
# avoid clean up if the previous test has been tear down
if self.case_number == 1 or self.case_number > 1000:
if self.case_number > 1000:
self.log.warn("TearDown for prev test failed. Will retry")
self.case_number -= 1000
self.tearDownEverything(reset_cluster_env_vars=False)
for cluster_name, cluster in self.cb_clusters.items():
if not self.skip_cluster_reset:
self.initialize_cluster(
cluster_name,
cluster,
services=None,
services_mem_quota_percent=mem_quota_percent)
# Update initial service map for the master node
self.cluster_util.update_cluster_nodes_service_list(cluster)
# Set this unconditionally
RestConnection(cluster.master).set_internalSetting(
"magmaMinMemoryQuota", 256)
# Enable dp_version since we need collections enabled
if self.enable_dp:
tasks = []
for server in self.cluster.servers:
task = self.node_utils.async_enable_dp(server)
tasks.append(task)
for task in tasks:
self.task_manager.get_task_result(task)
# Enforce tls on nodes of all clusters
if self.use_https and self.enforce_tls:
for _, cluster in self.cb_clusters.items():
tasks = []
for node in cluster.servers:
task = self.node_utils.async_enable_tls(node)
tasks.append(task)
for task in tasks:
self.task_manager.get_task_result(task)
self.log.info(
"Validating if services obey tls only on servers {0}".
format(cluster.servers))
status = self.cluster_util.check_if_services_obey_tls(
cluster.servers)
if not status:
self.fail("Services did not honor enforce tls")
# Enforce IPv4 or IPv6 or both
if self.ipv4_only or self.ipv6_only:
for _, cluster in self.cb_clusters.items():
status, msg = self.cluster_util.enable_disable_ip_address_family_type(
cluster, True, self.ipv4_only, self.ipv6_only)
if not status:
self.fail(msg)
self.standard = self.input.param("standard", "pkcs8")
self.passphrase_type = self.input.param("passphrase_type",
"script")
self.encryption_type = self.input.param("encryption_type",
"aes256")
if self.multiple_ca:
for _, cluster in self.cb_clusters.items():
cluster.x509 = x509main(
host=cluster.master,
standard=self.standard,
encryption_type=self.encryption_type,
passphrase_type=self.passphrase_type)
self.generate_and_upload_cert(cluster.servers,
cluster.x509,
upload_root_certs=True,
upload_node_certs=True,
upload_client_certs=True)
payload = "name=cbadminbucket&roles=admin&password=password"
rest = RestConnection(cluster.master)
rest.add_set_builtin_user("cbadminbucket", payload)
for cluster_name, cluster in self.cb_clusters.items():
self.modify_cluster_settings(cluster)
# Track test start time only if we need system log validation
if self.validate_system_event_logs:
self.system_events.set_test_start_time()
self.log_setup_status("OnPremBaseTest", "finished")
self.__log("started")
except Exception as e:
traceback.print_exc()
self.task.shutdown(force=True)
self.fail(e)
finally:
# Track test start time only if we need system log validation
if self.validate_system_event_logs:
self.system_events.set_test_start_time()
self.log_setup_status("OnPremBaseTest", "finished")
def initialize_cluster(self,
cluster_name,
cluster,
services=None,
services_mem_quota_percent=None):
self.log.info("Initializing cluster : {0}".format(cluster_name))
self.node_utils.reset_cluster_nodes(self.cluster_util, cluster)
if not services:
master_services = self.cluster_util.get_services(
cluster.servers[:1], self.services_init, start_node=0)
else:
master_services = self.cluster_util.get_services(
cluster.servers[:1], services, start_node=0)
if master_services is not None:
master_services = master_services[0].split(",")
self._initialize_nodes(
self.task,
cluster,
self.disabled_consistent_view,
self.rebalanceIndexWaitingDisabled,
self.rebalanceIndexPausingDisabled,
self.maxParallelIndexers,
self.maxParallelReplicaIndexers,
self.port,
self.quota_percent,
services=master_services,
services_mem_quota_percent=services_mem_quota_percent)
self.cluster_util.change_env_variables(cluster)
self.cluster_util.change_checkpoint_params(cluster)
self.log.info("Cluster %s initialized" % cluster_name)
def modify_cluster_settings(self, cluster):
if self.log_info:
self.cluster_util.change_log_info(cluster, self.log_info)
if self.log_location:
self.cluster_util.change_log_location(cluster, self.log_location)
if self.stat_info:
self.cluster_util.change_stat_info(cluster, self.stat_info)
if self.port_info:
self.cluster_util.change_port_info(cluster, self.port_info)
if self.port:
self.port = str(self.port)
def start_fetch_pcaps(self):
log_path = TestInputSingleton.input.param("logs_folder", "/tmp")
is_test_failed = self.is_test_failed()
self.node_utils.start_fetch_pcaps(self.servers, log_path,
is_test_failed)
def tearDown(self):
# Perform system event log validation and get failures (if any)
sys_event_validation_failure = None
if self.validate_system_event_logs:
sys_event_validation_failure = \
self.system_events.validate(self.cluster.master)
if self.ipv4_only or self.ipv6_only:
for _, cluster in self.cb_clusters.items():
self.cluster_util.enable_disable_ip_address_family_type(
cluster, False, self.ipv4_only, self.ipv6_only)
# Disable n2n encryption on nodes of all clusters
if self.use_https and self.enforce_tls:
for _, cluster in self.cb_clusters.items():
tasks = []
for node in cluster.servers:
task = self.node_utils.async_disable_tls(node)
tasks.append(task)
for task in tasks:
self.task_manager.get_task_result(task)
if self.multiple_ca:
CbServer.use_https = False
for _, cluster in self.cb_clusters.items():
rest = RestConnection(cluster.master)
rest.delete_builtin_user("cbadminbucket")
x509 = x509main(host=cluster.master)
x509.teardown_certs(servers=cluster.servers)
if self.sdk_client_pool:
self.sdk_client_pool.shutdown()
if self.collect_pcaps:
self.log.info("Starting Pcaps collection!!")
self.start_fetch_pcaps()
result = self.check_coredump_exist(self.servers, force_collect=True)
if self.skip_teardown_cleanup:
self.log.debug("Skipping tearDownEverything")
else:
self.tearDownEverything()
if not self.crash_warning:
self.assertFalse(result, msg="Cb_log file validation failed")
if self.crash_warning and result:
self.log.warn("CRASH | CRITICAL | WARN messages found in cb_logs")
# Fail test in case of sys_event_logging failure
if (not self.is_test_failed()) and sys_event_validation_failure:
self.fail(sys_event_validation_failure)
elif sys_event_validation_failure:
self.log.critical("System event log validation failed: %s" %
sys_event_validation_failure)
self.shutdown_task_manager()
def tearDownEverything(self, reset_cluster_env_vars=True):
for _, cluster in self.cb_clusters.items():
try:
test_failed = self.is_test_failed()
if test_failed:
# Collect logs because we have not shut things down
if self.get_cbcollect_info:
self.fetch_cb_collect_logs()
get_trace = \
TestInputSingleton.input.param("get_trace", None)
if get_trace:
tasks = []
for server in cluster.servers:
task = self.node_utils.async_get_trace(
server, get_trace)
tasks.append(task)
for task in tasks:
self.task_manager.get_task_result(task)
else:
self.log.critical("Skipping get_trace !!")
rest = RestConnection(cluster.master)
alerts = rest.get_alerts()
if alerts:
self.log.warn("Alerts found: {0}".format(alerts))
except BaseException as e:
# kill memcached
traceback.print_exc()
self.log.warning("Killing memcached due to {0}".format(e))
self.cluster_util.kill_memcached(cluster)
# Increase case_number to retry tearDown in setup for next test
self.case_number += 1000
finally:
if reset_cluster_env_vars:
self.cluster_util.reset_env_variables(cluster)
self.infra_log.info("========== tasks in thread pool ==========")
self.task_manager.print_tasks_in_pool()
self.infra_log.info("==========================================")
def __log(self, status):
try:
msg = "{0}: {1} {2}" \
.format(datetime.now(), self._testMethodName, status)
RestConnection(self.servers[0]).log_client_error(msg)
except Exception as e:
self.log.warning("Exception during REST log_client_error: %s" % e)
def _initialize_nodes(self,
task,
cluster,
disabled_consistent_view=None,
rebalance_index_waiting_disabled=None,
rebalance_index_pausing_disabled=None,
max_parallel_indexers=None,
max_parallel_replica_indexers=None,
port=None,
quota_percent=None,
services=None,
services_mem_quota_percent=None):
quota = 0
init_tasks = []
ssh_sessions = dict()
# Open ssh_connections for command execution
for server in cluster.servers:
ssh_sessions[server.ip] = RemoteMachineShellConnection(server)
for server in cluster.servers:
# Make sure that data_and index_path are writable by couchbase user
if not server.index_path:
server.index_path = server.data_path
if not server.cbas_path:
server.cbas_path = str([server.data_path])
if not server.eventing_path:
server.eventing_path = server.data_path
for path in set(
[_f for _f in [server.data_path, server.index_path] if _f]):
for cmd in ("rm -rf {0}/*".format(path),
"chown -R couchbase:couchbase {0}".format(path)):
ssh_sessions[server.ip].execute_command(cmd)
rest = RestConnection(server)
rest.set_data_path(data_path=server.data_path,
index_path=server.index_path,
cbas_path=server.cbas_path)
if cluster.master != server:
continue
init_port = port or server.port or '8091'
assigned_services = services
init_tasks.append(
task.async_init_node(
server,
disabled_consistent_view,
rebalance_index_waiting_disabled,
rebalance_index_pausing_disabled,
max_parallel_indexers,
max_parallel_replica_indexers,
init_port,
quota_percent,
services=assigned_services,
gsi_type=self.gsi_type,
services_mem_quota_percent=services_mem_quota_percent))
for _task in init_tasks:
node_quota = self.task_manager.get_task_result(_task)
if node_quota < quota or quota == 0:
quota = node_quota
if quota < 100 and not len(set([server.ip
for server in self.servers])) == 1:
self.log.warn("RAM quota was defined less than 100 MB:")
for server in cluster.servers:
ram = ssh_sessions[server.ip].extract_remote_info().ram
self.log.debug("Node: {0}: RAM: {1}".format(server.ip, ram))
# Close all ssh_connections
for server in cluster.servers:
ssh_sessions[server.ip].disconnect()
if self.jre_path:
for server in cluster.servers:
rest = RestConnection(server)
rest.set_jre_path(self.jre_path)
return quota
def fetch_cb_collect_logs(self):
log_path = TestInputSingleton.input.param("logs_folder", "/tmp")
is_single_node_server = len(self.servers) == 1
for _, cluster in self.cb_clusters.items():
rest = RestConnection(cluster.master)
nodes = rest.get_nodes(inactive=True)
# Creating cluster_util object to handle multi_cluster scenario
status = self.cluster_util.trigger_cb_collect_on_cluster(
rest, nodes, is_single_node_server)
if status is True:
self.cluster_util.wait_for_cb_collect_to_complete(rest)
self.cluster_util.copy_cb_collect_logs(rest, nodes, cluster,
log_path)
else:
self.log.error("API perform_cb_collect returned False")
def check_coredump_exist(self, servers, force_collect=False):
bin_cb = "/opt/couchbase/bin/"
lib_cb = "/opt/couchbase/var/lib/couchbase/"
# crash_dir = "/opt/couchbase/var/lib/couchbase/"
crash_dir_win = "c://CrashDumps"
result = False
self.data_sets = dict()
def find_index_of(str_list, sub_string):
for i in range(len(str_list)):
if sub_string in str_list[i]:
return i
return -1
def get_gdb(gdb_shell, dmp_path, dmp_name):
dmp_file = dmp_path + dmp_name
core_file = dmp_path + dmp_name.strip(".dmp") + ".core"
gdb_shell.execute_command("rm -rf " + core_file)
core_cmd = "/" + bin_cb + "minidump-2-core " + dmp_file + " > " + core_file
print("running: %s" % core_cmd)
gdb_shell.execute_command(core_cmd)
gdb = "gdb --batch {} -c {} -ex \"bt full\" -ex quit"\
.format(os.path.join(bin_cb, "memcached"), core_file)
print("running: %s" % gdb)
result = gdb_shell.execute_command(gdb)[0]
t_index = find_index_of(result, "Core was generated by")
result = result[t_index:]
result = " ".join(result)
return result
def get_full_thread_dump(shell):
cmd = 'gdb -p `(pidof memcached)` -ex "thread apply all bt" -ex detach -ex quit'
print("running: %s" % cmd)
thread_dump = shell.execute_command(cmd)[0]
index = find_index_of(
thread_dump, "Thread debugging using libthread_db enabled")
result = " ".join(thread_dump[index:])
return result
def run_cbanalyze_core(shell, core_file):
cbanalyze_core = os.path.join(bin_cb, "tools/cbanalyze-core")
cmd = '%s %s' % (cbanalyze_core, core_file)
print("running: %s" % cmd)
shell.execute_command(cmd)[0]
cbanalyze_log = core_file + ".log"
if shell.file_exists(os.path.dirname(cbanalyze_log),
os.path.basename(cbanalyze_log)):
log_path = TestInputSingleton.input.param(
"logs_folder", "/tmp")
shell.get_file(os.path.dirname(cbanalyze_log),
os.path.basename(cbanalyze_log), log_path)
def check_logs(grep_output_list):
"""
Check the grep's last line for the latest timestamp.
If this timestamp < start_timestamp of the test,
then return False (as the grep's output is from previous tests)
Note: This method works only if slave's time(timezone) matches
that of VM's. Else it won't be possible to compare timestamps
"""
last_line = grep_output_list[-1]
# eg: 2021-07-12T04:03:45
timestamp_regex = re.compile(
r"\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}")
match_obj = timestamp_regex.search(last_line)
if not match_obj:
self.log.critical("%s does not match any timestamp" %
last_line)
return True
timestamp = match_obj.group()
timestamp = datetime.strptime(timestamp, "%Y-%m-%dT%H:%M:%S")
self.log.info("Comparing timestamps: Log's latest timestamp: %s, "
"Test's start timestamp is %s" %
(timestamp, self.start_timestamp))
if timestamp > self.start_timestamp:
return True
else:
return False
for idx, server in enumerate(servers):
shell = RemoteMachineShellConnection(server)
shell.extract_remote_info()
crash_dir = lib_cb + "crash/"
if shell.info.type.lower() == "windows":
crash_dir = crash_dir_win
if int(server.port) in range(ClusterRun.port,
ClusterRun.port + 10):
crash_dir = os.path.join(
TestInputSingleton.input.servers[0].cli_path, "ns_server",
"data", "n_%s" % str(idx), "crash")
dmp_files = shell.execute_command("ls -lt " + crash_dir)[0]
dmp_files = [f for f in dmp_files if ".core" not in f]
dmp_files = [f for f in dmp_files if "total" not in f]
dmp_files = [f.split()[-1] for f in dmp_files if ".core" not in f]
dmp_files = [f.strip("\n") for f in dmp_files]
if dmp_files:
print("#" * 30)
print("%s: %d core dump seen" % (server.ip, len(dmp_files)))
print("%s: Stack Trace of first crash - %s\n%s" %
(server.ip, dmp_files[-1],
get_gdb(shell, crash_dir, dmp_files[-1])))
print("#" * 30)
result = get_full_thread_dump(shell)
print(result)
# print("#"*30)
# run_cbanalyze_core(shell, crash_dir + dmp_files[-1].strip(".dmp") + ".core")
# print("#"*30)
if self.stop_server_on_crash:
shell.stop_couchbase()
result = True
else:
self.log.debug(server.ip + ": No crash files found")
logs_dir = lib_cb + "logs/"
if int(server.port) in range(ClusterRun.port,
ClusterRun.port + 10):
logs_dir = os.path.join(
TestInputSingleton.input.servers[0].cli_path, "ns_server",
"logs", "n_%s" % str(idx))
# Perform log file searching based on the input yaml config
yaml = YAML()
with open("lib/couchbase_helper/error_log_config.yaml", "r") as fp:
y_data = yaml.load(fp.read())
for file_data in y_data["file_name_patterns"]:
log_files = shell.execute_command(
"ls " + os.path.join(logs_dir, file_data['file']))[0]
if len(log_files) == 0:
self.log.debug("%s: No '%s' files found" %
(server.ip, file_data['file']))
continue
if 'target_file_index' in file_data:
log_files = [
log_files[int(file_data['target_file_index'])]
]
for log_file in log_files:
log_file = log_file.strip("\n")
for grep_pattern in file_data['grep_for']:
grep_for_str = grep_pattern['string']
err_pattern = exclude_pattern = None
if 'error_patterns' in grep_pattern:
err_pattern = grep_pattern['error_patterns']
if 'exclude_patterns' in grep_pattern:
exclude_pattern = grep_pattern['exclude_patterns']
cmd_to_run = "grep -r '%s' %s" \
% (grep_for_str, log_file)
if exclude_pattern is not None:
for pattern in exclude_pattern:
cmd_to_run += " | grep -v '%s'" % pattern
grep_output = shell.execute_command(cmd_to_run)[0]
if grep_output and check_logs(grep_output):
regex = r"(\bkvstore-\d+)"
grep_str = "".join(grep_output)
kvstores = list(set(re.findall(regex, grep_str)))
self.data_sets[server] = kvstores
grep_str = None
if err_pattern is not None:
for pattern in err_pattern:
index = find_index_of(grep_output, pattern)
grep_output = grep_output[:index]
if grep_output:
self.log.info("unwanted messages in %s" %
log_file)
if check_logs(grep_output):
self.log.critical(
"%s: Found '%s' logs - %s" %
(server.ip, grep_for_str,
"".join(grep_output)))
result = True
break
else:
if grep_output \
and check_logs(grep_output):
self.log.info("unwanted messages in %s" %
log_file)
self.log.critical(
"%s: Found '%s' logs - %s" %
(server.ip, grep_for_str, grep_output))
result = True
break
if result is True:
if self.stop_server_on_crash:
shell.stop_couchbase()
break
shell.disconnect()
if result and force_collect and not self.stop_server_on_crash:
self.fetch_cb_collect_logs()
self.get_cbcollect_info = False
if (self.is_test_failed() or result) and self.collect_data:
self.copy_data_on_slave()
return result
def copy_data_on_slave(self, servers=None):
log_path = TestInputSingleton.input.param("logs_folder", "/tmp")
if servers is None:
servers = self.cluster.nodes_in_cluster
for node in servers:
if "kv" not in node.services.lower():
servers.remove(node)
if type(servers) is not list:
servers = [servers]
remote_path = RestConnection(servers[0]).get_data_path()
file_path = os.path.join(remote_path, self.cluster.buckets[0].name)
file_name = self.cluster.buckets[0].name + ".tar.gz"
def get_tar(remotepath, filepath, filename, servers, todir="."):
if type(servers) is not list:
servers = [servers]
for server in servers:
shell = RemoteMachineShellConnection(server)
_ = shell.execute_command("tar -zcvf %s.tar.gz %s" %
(filepath, filepath))
file_check = shell.file_exists(remotepath, filename)
if not file_check:
self.log.error(
"Tar File {} doesn't exist".format(filename))
tar_file_copied = shell.get_file(remotepath, filename, todir)
if not tar_file_copied:
self.log.error("Failed to copy Tar file")
_ = shell.execute_command("rm -rf %s.tar.gz" % filepath)
copy_path_msg_format = "Copying data, Server :: %s, Path :: %s"
'''
Temporarily enabling data copy
of all nodes irrespective of nodes in
data_sets
'''
if False and self.data_sets and self.bucket_storage == "magma":
self.log.critical("data_sets ==> {}".format(self.data_sets))
wal_tar = "wal.tar.gz"
config_json_tar = "config.json.tar.gz"
for server, kvstores in self.data_sets.items():
shell = RemoteMachineShellConnection(server)
if not kvstores:
copy_to_path = os.path.join(log_path,
server.ip.replace(".", "_"))
if not os.path.isdir(copy_to_path):
os.makedirs(copy_to_path, 0o777)
self.log.info(copy_path_msg_format %
(server.ip, copy_to_path))
get_tar(remote_path,
file_path,
file_name,
server,
todir=copy_to_path)
else:
for kvstore in kvstores:
if int(kvstore.split("-")[1]) >= self.vbuckets:
continue
kvstore_path = shell.execute_command(
"find %s -type d -name '%s'" %
(remote_path, kvstore))[0][0]
magma_dir = kvstore_path.split(kvstore)[0]
wal_path = kvstore_path.split(kvstore)[0] + "wal"
config_json_path = kvstore_path.split(kvstore)[0]
+ "config.json"
kvstore_path = kvstore_path.split(kvstore)[0] + kvstore
kvstore_tar = kvstore + ".tar.gz"
copy_to_path = os.path.join(log_path, kvstore)
if not os.path.isdir(copy_to_path):
os.makedirs(copy_to_path, 0o777)
self.log.info(copy_path_msg_format %
(server.ip, copy_to_path))
get_tar(magma_dir,
kvstore_path,
kvstore_tar,
server,
todir=copy_to_path)
get_tar(magma_dir,
wal_path,
wal_tar,
server,
todir=copy_to_path)
get_tar(magma_dir,
config_json_path,
config_json_tar,
server,
todir=copy_to_path)
else:
for server in servers:
copy_to_path = os.path.join(log_path,
server.ip.replace(".", "_"))
if not os.path.isdir(copy_to_path):
os.makedirs(copy_to_path, 0o777)
self.log.info(copy_path_msg_format % (server.ip, copy_to_path))
get_tar(remote_path,
file_path,
file_name,
server,
todir=copy_to_path)
def generate_and_upload_cert(self,
servers,
x509,
generate_certs=True,
delete_inbox_folder=True,
upload_root_certs=True,
upload_node_certs=True,
delete_out_of_the_box_CAs=True,
upload_client_certs=True):
if generate_certs:
x509.generate_multiple_x509_certs(servers=servers)
if delete_inbox_folder:
for server in servers:
x509.delete_inbox_folder_on_server(server=server)
if upload_root_certs:
for server in servers:
_ = x509.upload_root_certs(server)
if upload_node_certs:
x509.upload_node_certs(servers=servers)
if delete_out_of_the_box_CAs:
for node in servers:
x509.delete_unused_out_of_the_box_CAs(server=node)
if upload_client_certs:
x509.upload_client_cert_settings(server=servers[0])
def setUp(self):
self.log = logging.getLogger()
self.tear_down_while_setup = True
self.input = TestInputSingleton.input
self.primary_index_created = False
self.sdk_client_type = self.input.param("sdk_client_type", "java")
if self.input.param("log_level", None):
self.log.setLevel(level=0)
for hd in self.log.handlers:
if str(hd.__class__).find('FileHandler') != -1:
hd.setLevel(level=logging.DEBUG)
else:
hd.setLevel(level=getattr(
logging, self.input.param("log_level", None)))
self.servers = self.input.servers
self.buckets = []
self.case_number = self.input.param("case_number", 0)
self.thread_to_use = self.input.param("threads_to_use", 10)
self.cluster = CBCluster(servers=self.input.servers)
self.task_manager = TaskManager(self.thread_to_use)
self.cluster_util = ClusterUtils(self.cluster, self.task_manager)
self.task = ServerTasks(self.task_manager)
self.bucket_util = BucketUtils(self.cluster, self.cluster_util,
self.task)
self.cleanup = False
self.nonroot = False
self.test_failures = list()
shell = RemoteMachineShellConnection(self.cluster.master)
self.os_info = shell.extract_remote_info().type.lower()
if self.os_info != 'windows':
if self.cluster.master.ssh_username != "root":
self.nonroot = True
shell.disconnect()
""" some tests need to bypass checking cb server at set up
to run installation """
self.skip_init_check_cbserver = self.input.param(
"skip_init_check_cbserver", False)
try:
# Framework specific params
self.skip_setup_cleanup = self.input.param("skip_setup_cleanup",
False)
self.log_info = self.input.param("log_info", None)
self.log_location = self.input.param("log_location", None)
# kill hang test and jump to next one.
self.test_timeout = self.input.param("test_timeout", 3600)
# Bucket specific params
self.bucket_type = self.input.param("bucket_type", "membase")
self.bucket_size = self.input.param("bucket_size", None)
self.standard_buckets = self.input.param("standard_buckets", 1)
self.vbuckets = self.input.param("vbuckets", 1024)
self.num_replicas = self.input.param("replicas", 1)
self.active_resident_threshold = int(
self.input.param("active_resident_threshold", 100))
self.compression_mode = self.input.param("compression_mode",
'passive')
# end of bucket parameters spot (this is ongoing)
# Doc specific params
self.key_size = self.input.param("key_size", 0)
self.doc_size = self.input.param("doc_size", 10)
self.doc_type = self.input.param("doc_type", "json")
self.num_items = self.input.param("num_items", 10000)
self.target_vbucket = self.input.param("target_vbucket", None)
self.maxttl = self.input.param("maxttl", 0)
self.transaction_timeout = self.input.param(
"transaction_timeout", 100)
self.transaction_commit = self.input.param("transaction_commit",
True)
#transactions param
self.transaction_timeout = self.input.param(
"transaction_timeout", 100)
self.transaction_commit = self.input.param("transaction_commit",
True)
self.update_count = self.input.param("update_count", 1)
self.sync = self.input.param("sync", True)
self.default_bucket = self.input.param("default_bucket", True)
self.num_buckets = self.input.param("num_buckets", 0)
# Client specific params
self.sdk_compression = self.input.param("sdk_compression", True)
self.replicate_to = self.input.param("replicate_to", 0)
self.persist_to = self.input.param("persist_to", 0)
self.sdk_retries = self.input.param("sdk_retries", 5)
self.sdk_timeout = self.input.param("sdk_timeout", 5)
self.durability_level = self.input.param("durability", "")
self.durability_timeout = self.input.param("durability_timeout", 0)
self.index_quota_percent = self.input.param(
"index_quota_percent", None)
self.num_servers = self.input.param("servers",
len(self.cluster.servers))
# initial number of items in the cluster
self.services_init = self.input.param("services_init", None)
self.nodes_init = self.input.param("nodes_init", 1)
self.nodes_in = self.input.param("nodes_in", 1)
self.nodes_out = self.input.param("nodes_out", 1)
self.services_in = self.input.param("services_in", None)
self.forceEject = self.input.param("forceEject", False)
self.value_size = self.input.param("value_size", 1)
self.wait_timeout = self.input.param("wait_timeout", 60)
self.dgm_run = self.input.param("dgm_run", False)
self.verify_unacked_bytes = self.input.param(
"verify_unacked_bytes", False)
self.disabled_consistent_view = self.input.param(
"disabled_consistent_view", None)
self.rebalanceIndexWaitingDisabled = self.input.param(
"rebalanceIndexWaitingDisabled", None)
self.rebalanceIndexPausingDisabled = self.input.param(
"rebalanceIndexPausingDisabled", None)
self.maxParallelIndexers = self.input.param(
"maxParallelIndexers", None)
self.maxParallelReplicaIndexers = self.input.param(
"maxParallelReplicaIndexers", None)
self.quota_percent = self.input.param("quota_percent", None)
self.port = None
self.stat_info = self.input.param("stat_info", None)
self.port_info = self.input.param("port_info", None)
if not hasattr(self, 'skip_buckets_handle'):
self.skip_buckets_handle = self.input.param(
"skip_buckets_handle", False)
self.gsi_type = self.input.param("gsi_type", 'plasma')
# jre-path for cbas
self.jre_path = self.input.param("jre_path", None)
if self.skip_setup_cleanup:
self.buckets = self.bucket_util.get_all_buckets()
return
if not self.skip_init_check_cbserver:
self.cb_version = None
if RestHelper(RestConnection(
self.cluster.master)).is_ns_server_running():
"""
Since every new couchbase version, there will be new
features that test code won't work on previous release.
So we need to get couchbase version to filter out
those tests.
"""
self.cb_version = RestConnection(
self.cluster.master).get_nodes_version()
else:
self.log.info("couchbase server does not run yet")
self.protocol = self.cluster_util.get_protocol_type()
self.services_map = None
self.__log_setup_status("started")
if not self.skip_buckets_handle and not self.skip_init_check_cbserver:
self.cluster_util.cluster_cleanup(self.bucket_util)
# avoid any cluster operations in setup for new upgrade
# & upgradeXDCR tests
if str(self.__class__).find('newupgradetests') != -1 or \
str(self.__class__).find('upgradeXDCR') != -1 or \
str(self.__class__).find('Upgrade_EpTests') != -1 or \
hasattr(self, 'skip_buckets_handle') and \
self.skip_buckets_handle:
self.log.info("any cluster operation in setup will be skipped")
self.primary_index_created = True
self.__log_setup_status("finished")
return
# avoid clean up if the previous test has been tear down
if self.case_number == 1 or self.case_number > 1000:
if self.case_number > 1000:
self.log.warn(
"TearDown for previous test failed. will retry..")
self.case_number -= 1000
self.cleanup = True
if not self.skip_init_check_cbserver:
self.tearDownEverything()
self.tear_down_while_setup = False
if not self.skip_init_check_cbserver:
self.log.info("Initializing cluster")
# self.cluster_util.reset_cluster()
master_services = self.cluster_util.get_services(
self.servers[:1], self.services_init, start_node=0)
if master_services is not None:
master_services = master_services[0].split(",")
self.quota = self._initialize_nodes(
self.task,
self.cluster.servers,
self.disabled_consistent_view,
self.rebalanceIndexWaitingDisabled,
self.rebalanceIndexPausingDisabled,
self.maxParallelIndexers,
self.maxParallelReplicaIndexers,
self.port,
self.quota_percent,
services=master_services)
self.cluster_util.change_env_variables()
self.cluster_util.change_checkpoint_params()
self.log.info("Cluster initialized")
else:
self.quota = ""
if self.input.param("log_info", None):
self.cluster_util.change_log_info()
if self.input.param("log_location", None):
self.cluster_util.change_log_location()
if self.input.param("stat_info", None):
self.cluster_util.change_stat_info()
if self.input.param("port_info", None):
self.cluster_util.change_port_info()
if self.input.param("port", None):
self.port = str(self.input.param("port", None))
self.__log_setup_status("finished")
if not self.skip_init_check_cbserver:
self.__log("started")
self.sleep(5)
except Exception, e:
traceback.print_exc()
self.task.shutdown(force=True)
self.fail(e)
class BaseTestCase(unittest.TestCase):
def setUp(self):
self.input = TestInputSingleton.input
# Framework specific parameters
self.log_level = self.input.param("log_level", "info").upper()
self.infra_log_level = self.input.param("infra_log_level",
"error").upper()
self.skip_setup_cleanup = self.input.param("skip_setup_cleanup", False)
self.tear_down_while_setup = self.input.param("tear_down_while_setup",
True)
self.test_timeout = self.input.param("test_timeout", 3600)
self.thread_to_use = self.input.param("threads_to_use", 30)
self.case_number = self.input.param("case_number", 0)
# End of framework parameters
# Cluster level info settings
self.log_info = self.input.param("log_info", None)
self.log_location = self.input.param("log_location", None)
self.stat_info = self.input.param("stat_info", None)
self.port = self.input.param("port", None)
self.port_info = self.input.param("port_info", None)
self.servers = self.input.servers
self.cb_clusters = OrderedDict()
self.num_servers = self.input.param("servers", len(self.servers))
self.vbuckets = self.input.param("vbuckets", CbServer.total_vbuckets)
self.primary_index_created = False
self.index_quota_percent = self.input.param("index_quota_percent",
None)
self.gsi_type = self.input.param("gsi_type", 'plasma')
# CBAS setting
self.jre_path = self.input.param("jre_path", None)
self.enable_dp = self.input.param("enable_dp", False)
# End of cluster info parameters
# Bucket specific params
self.bucket_type = self.input.param("bucket_type", Bucket.Type.MEMBASE)
self.bucket_ttl = self.input.param("bucket_ttl", 0)
self.bucket_size = self.input.param("bucket_size", None)
self.bucket_conflict_resolution_type = \
self.input.param("bucket_conflict_resolution",
Bucket.ConflictResolution.SEQ_NO)
self.bucket_replica_index = self.input.param("bucket_replica_index", 1)
self.bucket_eviction_policy = \
self.input.param("bucket_eviction_policy",
Bucket.EvictionPolicy.VALUE_ONLY)
self.flush_enabled = self.input.param("flushEnabled",
Bucket.FlushBucket.DISABLED)
self.bucket_time_sync = self.input.param("bucket_time_sync", False)
self.standard_buckets = self.input.param("standard_buckets", 1)
self.num_replicas = self.input.param("replicas", Bucket.ReplicaNum.ONE)
self.active_resident_threshold = \
int(self.input.param("active_resident_threshold", 100))
self.compression_mode = \
self.input.param("compression_mode",
Bucket.CompressionMode.PASSIVE)
self.bucket_storage = \
self.input.param("bucket_storage",
Bucket.StorageBackend.couchstore)
if self.bucket_storage == Bucket.StorageBackend.magma:
self.bucket_eviction_policy = Bucket.EvictionPolicy.FULL_EVICTION
self.scope_name = self.input.param("scope", CbServer.default_scope)
self.collection_name = self.input.param("collection",
CbServer.default_collection)
self.bucket_durability_level = self.input.param(
"bucket_durability", Bucket.DurabilityLevel.NONE).upper()
self.bucket_purge_interval = self.input.param("bucket_purge_interval",
1)
self.bucket_durability_level = \
BucketDurability[self.bucket_durability_level]
# End of bucket parameters
# Doc specific params
self.key = self.input.param("key", "test_docs")
self.key_size = self.input.param("key_size", 8)
self.doc_size = self.input.param("doc_size", 256)
self.sub_doc_size = self.input.param("sub_doc_size", 10)
self.doc_type = self.input.param("doc_type", "json")
self.num_items = self.input.param("num_items", 100000)
self.target_vbucket = self.input.param("target_vbucket", None)
self.maxttl = self.input.param("maxttl", 0)
self.random_exp = self.input.param("random_exp", False)
self.randomize_doc_size = self.input.param("randomize_doc_size", False)
self.randomize_value = self.input.param("randomize_value", False)
self.rev_write = self.input.param("rev_write", False)
self.rev_read = self.input.param("rev_read", False)
self.rev_update = self.input.param("rev_update", False)
self.rev_del = self.input.param("rev_del", False)
self.random_key = self.input.param("random_key", False)
self.mix_key_size = self.input.param("mix_key_size", False)
# End of doc specific parameters
# Transactions parameters
self.transaction_timeout = self.input.param("transaction_timeout", 100)
self.transaction_commit = self.input.param("transaction_commit", True)
self.update_count = self.input.param("update_count", 1)
self.sync = self.input.param("sync", True)
self.default_bucket = self.input.param("default_bucket", True)
self.num_buckets = self.input.param("num_buckets", 0)
self.atomicity = self.input.param("atomicity", False)
self.defer = self.input.param("defer", False)
# end of transaction parameters
# Client specific params
self.sdk_client_type = self.input.param("sdk_client_type", "java")
self.replicate_to = self.input.param("replicate_to", 0)
self.persist_to = self.input.param("persist_to", 0)
self.sdk_retries = self.input.param("sdk_retries", 5)
self.sdk_timeout = self.input.param("sdk_timeout", 5)
self.time_unit = self.input.param("time_unit", "seconds")
self.durability_level = self.input.param("durability", "").upper()
self.sdk_client_pool = self.input.param("sdk_client_pool", None)
self.sdk_pool_capacity = self.input.param("sdk_pool_capacity", 1)
# Client compression settings
self.sdk_compression = self.input.param("sdk_compression", None)
compression_min_ratio = self.input.param("min_ratio", None)
compression_min_size = self.input.param("min_size", None)
if type(self.sdk_compression) is bool:
self.sdk_compression = {"enabled": self.sdk_compression}
if compression_min_size:
self.sdk_compression["minSize"] = compression_min_size
if compression_min_ratio:
self.sdk_compression["minRatio"] = compression_min_ratio
# Doc Loader Params
self.process_concurrency = self.input.param("process_concurrency", 20)
self.batch_size = self.input.param("batch_size", 2000)
self.dgm_batch = self.input.param("dgm_batch", 5000)
self.ryow = self.input.param("ryow", False)
self.check_persistence = self.input.param("check_persistence", False)
# End of client specific parameters
# initial number of items in the cluster
self.services_init = self.input.param("services_init", None)
self.nodes_init = self.input.param("nodes_init", 1)
self.nodes_in = self.input.param("nodes_in", 1)
self.nodes_out = self.input.param("nodes_out", 1)
self.services_in = self.input.param("services_in", None)
self.forceEject = self.input.param("forceEject", False)
self.wait_timeout = self.input.param("wait_timeout", 120)
self.verify_unacked_bytes = \
self.input.param("verify_unacked_bytes", False)
self.disabled_consistent_view = \
self.input.param("disabled_consistent_view", None)
self.rebalanceIndexWaitingDisabled = \
self.input.param("rebalanceIndexWaitingDisabled", None)
self.rebalanceIndexPausingDisabled = \
self.input.param("rebalanceIndexPausingDisabled", None)
self.maxParallelIndexers = \
self.input.param("maxParallelIndexers", None)
self.maxParallelReplicaIndexers = \
self.input.param("maxParallelReplicaIndexers", None)
self.quota_percent = self.input.param("quota_percent", 90)
self.skip_buckets_handle = self.input.param("skip_buckets_handle",
False)
# SDKClientPool object for creating generic clients across tasks
if self.sdk_client_pool is True:
self.init_sdk_pool_object()
# Initiate logging variables
self.log = logger.get("test")
self.infra_log = logger.get("infra")
self.cleanup_pcaps()
self.collect_pcaps = self.input.param("collect_pcaps", False)
if self.collect_pcaps:
self.start_collect_pcaps()
# variable for log collection using cbCollect
self.get_cbcollect_info = self.input.param("get-cbcollect-info", False)
# Variable for initializing the current (start of test) timestamp
self.start_timestamp = datetime.now()
'''
Be careful while using this flag.
This is only and only for stand-alone tests.
During bugs reproductions, when a crash is seen
stop_server_on_crash will stop the server
so that we can collect data/logs/dumps at the right time
'''
self.stop_server_on_crash = self.input.param("stop_server_on_crash",
False)
self.collect_data = self.input.param("collect_data", False)
# Configure loggers
self.log.setLevel(self.log_level)
self.infra_log.setLevel(self.infra_log_level)
# Support lib objects for testcase execution
self.task_manager = TaskManager(self.thread_to_use)
self.task = ServerTasks(self.task_manager)
# End of library object creation
self.sleep = sleep
self.cleanup = False
self.nonroot = False
self.test_failure = None
self.crash_warning = self.input.param("crash_warning", False)
self.summary = TestSummary(self.log)
# Populate memcached_port in case of cluster_run
cluster_run_base_port = ClusterRun.port
if int(self.input.servers[0].port) == ClusterRun.port:
for server in self.input.servers:
server.port = cluster_run_base_port
cluster_run_base_port += 1
# If not defined in node.ini under 'memcached_port' section
if server.memcached_port is CbServer.memcached_port:
server.memcached_port = \
ClusterRun.memcached_port \
+ (2 * (int(server.port) - ClusterRun.port))
self.log_setup_status(self.__class__.__name__, "started")
cluster_name_format = "C%s"
default_cluster_index = counter_index = 1
if len(self.input.clusters) > 1:
# Multi cluster setup
for _, nodes in self.input.clusters.iteritems():
cluster_name = cluster_name_format % counter_index
tem_cluster = CBCluster(name=cluster_name,
servers=nodes,
vbuckets=self.vbuckets)
self.cb_clusters[cluster_name] = tem_cluster
counter_index += 1
else:
# Single cluster
cluster_name = cluster_name_format % counter_index
self.cb_clusters[cluster_name] = CBCluster(name=cluster_name,
servers=self.servers,
vbuckets=self.vbuckets)
# Initialize self.cluster with first available cluster as default
self.cluster = self.cb_clusters[cluster_name_format %
default_cluster_index]
self.cluster_util = ClusterUtils(self.task_manager)
self.bucket_util = BucketUtils(self.cluster_util, self.task)
if self.standard_buckets > 10:
self.bucket_util.change_max_buckets(self.cluster.master,
self.standard_buckets)
for cluster_name, cluster in self.cb_clusters.items():
shell = RemoteMachineShellConnection(cluster.master)
self.os_info = shell.extract_remote_info().type.lower()
if self.os_info != 'windows':
if cluster.master.ssh_username != "root":
self.nonroot = True
shell.disconnect()
break
shell.disconnect()
""" some tests need to bypass checking cb server at set up
to run installation """
self.skip_init_check_cbserver = \
self.input.param("skip_init_check_cbserver", False)
try:
if self.skip_setup_cleanup:
self.cluster.buckets = self.bucket_util.get_all_buckets(
self.cluster)
return
self.services_map = None
self.log_setup_status("BaseTestCase", "started")
for cluster_name, cluster in self.cb_clusters.items():
if not self.skip_buckets_handle \
and not self.skip_init_check_cbserver:
self.log.debug("Cleaning up cluster")
self.cluster_util.cluster_cleanup(cluster,
self.bucket_util)
# Avoid cluster operations in setup for new upgrade / upgradeXDCR
if str(self.__class__).find('newupgradetests') != -1 or \
str(self.__class__).find('upgradeXDCR') != -1 or \
str(self.__class__).find('Upgrade_EpTests') != -1 or \
self.skip_buckets_handle:
self.log.warning("Cluster operation in setup will be skipped")
self.primary_index_created = True
self.log_setup_status("BaseTestCase", "finished")
return
# avoid clean up if the previous test has been tear down
if self.case_number == 1 or self.case_number > 1000:
if self.case_number > 1000:
self.log.warn("TearDown for prev test failed. Will retry")
self.case_number -= 1000
self.cleanup = True
if not self.skip_init_check_cbserver:
self.tearDownEverything()
self.tear_down_while_setup = False
if not self.skip_init_check_cbserver:
for cluster_name, cluster in self.cb_clusters.items():
self.initialize_cluster(cluster_name,
cluster,
services=None)
else:
self.quota = ""
# Enable dp_version since we need collections enabled
if self.enable_dp:
for server in self.cluster.servers:
shell_conn = RemoteMachineShellConnection(server)
cb_cli = CbCli(shell_conn)
cb_cli.enable_dp()
shell_conn.disconnect()
for cluster_name, cluster in self.cb_clusters.items():
self.modify_cluster_settings(cluster)
self.log_setup_status("BaseTestCase", "finished")
if not self.skip_init_check_cbserver:
self.__log("started")
except Exception as e:
traceback.print_exc()
self.task.shutdown(force=True)
self.fail(e)
def initialize_cluster(self, cluster_name, cluster, services=None):
self.log.info("Initializing cluster : {0}".format(cluster_name))
self.cluster_util.reset_cluster(cluster)
if not services:
master_services = self.cluster_util.get_services(
cluster.servers[:1], self.services_init, start_node=0)
else:
master_services = self.cluster_util.get_services(
cluster.servers[:1], services, start_node=0)
if master_services is not None:
master_services = master_services[0].split(",")
self.quota = self._initialize_nodes(self.task,
cluster,
self.disabled_consistent_view,
self.rebalanceIndexWaitingDisabled,
self.rebalanceIndexPausingDisabled,
self.maxParallelIndexers,
self.maxParallelReplicaIndexers,
self.port,
self.quota_percent,
services=master_services)
self.cluster_util.change_env_variables(cluster)
self.cluster_util.change_checkpoint_params(cluster)
self.log.info("Cluster %s initialized" % cluster_name)
def modify_cluster_settings(self, cluster):
if self.log_info:
self.cluster_util.change_log_info(cluster, self.log_info)
if self.log_location:
self.cluster_util.change_log_location(cluster, self.log_location)
if self.stat_info:
self.cluster_util.change_stat_info(cluster, self.stat_info)
if self.port_info:
self.cluster_util.change_port_info(cluster, self.port_info)
if self.port:
self.port = str(self.port)
def cleanup_pcaps(self):
for server in self.servers:
shell = RemoteMachineShellConnection(server)
# Stop old instances of tcpdump if still running
stop_tcp_cmd = "if [[ \"$(pgrep tcpdump)\" ]]; " \
"then kill -s TERM $(pgrep tcpdump); fi"
_, _ = shell.execute_command(stop_tcp_cmd)
shell.execute_command("rm -rf pcaps")
shell.execute_command("rm -rf " + server.ip + "_pcaps.zip")
shell.disconnect()
def start_collect_pcaps(self):
for server in self.servers:
shell = RemoteMachineShellConnection(server)
# Create path for storing pcaps
create_path = "mkdir -p pcaps"
o, e = shell.execute_command(create_path)
shell.log_command_output(o, e)
# Install tcpdump command if it doesn't exist
o, e = shell.execute_command("yum install -y tcpdump")
shell.log_command_output(o, e)
# Install screen command if it doesn't exist
o, e = shell.execute_command("yum install -y screen")
shell.log_command_output(o, e)
# Execute the tcpdump command
tcp_cmd = "screen -dmS test bash -c \"tcpdump -C 500 -W 10 " \
"-w pcaps/pack-dump-file.pcap -i eth0 -s 0 tcp\""
o, e = shell.execute_command(tcp_cmd)
shell.log_command_output(o, e)
shell.disconnect()
def start_fetch_pcaps(self):
log_path = TestInputSingleton.input.param("logs_folder", "/tmp")
for server in self.servers:
remote_client = RemoteMachineShellConnection(server)
# stop tcdump
stop_tcp_cmd = "if [[ \"$(pgrep tcpdump)\" ]]; " \
"then kill -s TERM $(pgrep tcpdump); fi"
o, e = remote_client.execute_command(stop_tcp_cmd)
remote_client.log_command_output(o, e)
if self.is_test_failed():
# install zip unzip
o, e = remote_client.execute_command(
"yum install -y zip unzip")
remote_client.log_command_output(o, e)
# zip the pcaps folder
zip_cmd = "zip -r " + server.ip + "_pcaps.zip pcaps"
o, e = remote_client.execute_command(zip_cmd)
remote_client.log_command_output(o, e)
# transfer the zip file
zip_file_copied = remote_client.get_file(
"/root", os.path.basename(server.ip + "_pcaps.zip"),
log_path)
self.log.info("%s node pcap zip copied on client : %s" %
(server.ip, zip_file_copied))
if zip_file_copied:
# Remove the zips
remote_client.execute_command("rm -rf " + server.ip +
"_pcaps.zip")
# Remove pcaps
remote_client.execute_command("rm -rf pcaps")
remote_client.disconnect()
def tearDown(self):
self.task_manager.shutdown_task_manager()
self.task.shutdown(force=True)
self.task_manager.abort_all_tasks()
if self.sdk_client_pool:
self.sdk_client_pool.shutdown()
if self.collect_pcaps:
self.log.info("Starting Pcaps collection!!")
self.start_fetch_pcaps()
result = self.check_coredump_exist(self.servers, force_collect=True)
self.tearDownEverything()
if not self.crash_warning:
self.assertFalse(result, msg="Cb_log file validation failed")
if self.crash_warning and result:
self.log.warn("CRASH | CRITICAL | WARN messages found in cb_logs")
def tearDownEverything(self):
if self.skip_setup_cleanup:
return
for _, cluster in self.cb_clusters.items():
try:
if self.skip_buckets_handle:
return
test_failed = self.is_test_failed()
if test_failed \
and TestInputSingleton.input.param("stop-on-failure",
False) \
or self.input.param("skip_cleanup", False):
self.log.warn("CLEANUP WAS SKIPPED")
else:
if test_failed:
# Collect logs because we have not shut things down
if self.get_cbcollect_info:
self.fetch_cb_collect_logs()
get_trace = \
TestInputSingleton.input.param("get_trace", None)
if get_trace:
for server in cluster.servers:
shell = \
RemoteMachineShellConnection(server)
output, _ = shell.execute_command(
"ps -aef|grep %s" % get_trace)
output = shell.execute_command(
"pstack %s" % output[0].split()[1].strip())
self.infra_log.debug(output[0])
shell.disconnect()
else:
self.log.critical("Skipping get_trace !!")
rest = RestConnection(cluster.master)
alerts = rest.get_alerts()
if alerts is not None and len(alerts) != 0:
self.infra_log.warn("Alerts found: {0}".format(alerts))
self.log.debug("Cleaning up cluster")
self.cluster_util.cluster_cleanup(cluster,
self.bucket_util)
except BaseException as e:
# kill memcached
traceback.print_exc()
self.log.warning("Killing memcached due to {0}".format(e))
self.cluster_util.kill_memcached(cluster)
# Increase case_number to retry tearDown in setup for next test
self.case_number += 1000
finally:
# stop all existing task manager threads
if self.cleanup:
self.cleanup = False
else:
self.cluster_util.reset_env_variables(cluster)
self.infra_log.info("========== tasks in thread pool ==========")
self.task_manager.print_tasks_in_pool()
self.infra_log.info("==========================================")
if not self.tear_down_while_setup:
self.task_manager.shutdown_task_manager()
self.task.shutdown(force=True)
def is_test_failed(self):
return (hasattr(self, '_resultForDoCleanups')
and len(self._resultForDoCleanups.failures
or self._resultForDoCleanups.errors)) \
or (hasattr(self, '_exc_info')
and self._exc_info()[1] is not None)
def handle_setup_exception(self, exception_obj):
# Shutdown client pool in case of any error before failing
if self.sdk_client_pool is not None:
self.sdk_client_pool.shutdown()
# print the tracback of the failure
traceback.print_exc()
# Throw the exception so that the test will fail at setUp
raise exception_obj
def __log(self, status):
try:
msg = "{0}: {1} {2}" \
.format(datetime.now(), self._testMethodName, status)
RestConnection(self.servers[0]).log_client_error(msg)
except Exception as e:
self.log.warning("Exception during REST log_client_error: %s" % e)
def log_setup_status(self, class_name, status, stage="setup"):
self.log.info("========= %s %s %s for test #%d %s =========" %
(class_name, stage, status, self.case_number,
self._testMethodName))
def _initialize_nodes(self,
task,
cluster,
disabled_consistent_view=None,
rebalance_index_waiting_disabled=None,
rebalance_index_pausing_disabled=None,
max_parallel_indexers=None,
max_parallel_replica_indexers=None,
port=None,
quota_percent=None,
services=None):
quota = 0
init_tasks = []
ssh_sessions = dict()
# Open ssh_connections for command execution
for server in cluster.servers:
ssh_sessions[server.ip] = RemoteMachineShellConnection(server)
for server in cluster.servers:
# Make sure that data_and index_path are writable by couchbase user
if not server.index_path:
server.index_path = server.data_path
for path in set(
[_f for _f in [server.data_path, server.index_path] if _f]):
for cmd in ("rm -rf {0}/*".format(path),
"chown -R couchbase:couchbase {0}".format(path)):
ssh_sessions[server.ip].execute_command(cmd)
rest = RestConnection(server)
rest.set_data_path(data_path=server.data_path,
index_path=server.index_path,
cbas_path=server.cbas_path)
init_port = port or server.port or '8091'
assigned_services = services
if cluster.master != server:
assigned_services = None
init_tasks.append(
task.async_init_node(
server,
disabled_consistent_view,
rebalance_index_waiting_disabled,
rebalance_index_pausing_disabled,
max_parallel_indexers,
max_parallel_replica_indexers,
init_port,
quota_percent,
services=assigned_services,
index_quota_percent=self.index_quota_percent,
gsi_type=self.gsi_type))
for _task in init_tasks:
node_quota = self.task_manager.get_task_result(_task)
if node_quota < quota or quota == 0:
quota = node_quota
if quota < 100 and not len(set([server.ip
for server in self.servers])) == 1:
self.log.warn("RAM quota was defined less than 100 MB:")
for server in cluster.servers:
ram = ssh_sessions[server.ip].extract_remote_info().ram
self.log.debug("Node: {0}: RAM: {1}".format(server.ip, ram))
# Close all ssh_connections
for server in cluster.servers:
ssh_sessions[server.ip].disconnect()
if self.jre_path:
for server in cluster.servers:
rest = RestConnection(server)
rest.set_jre_path(self.jre_path)
return quota
def fetch_cb_collect_logs(self):
log_path = TestInputSingleton.input.param("logs_folder", "/tmp")
is_single_node_server = len(self.servers) == 1
for _, cluster in self.cb_clusters.items():
rest = RestConnection(cluster.master)
nodes = rest.get_nodes()
# Creating cluster_util object to handle multi_cluster scenario
status = self.cluster_util.trigger_cb_collect_on_cluster(
rest, nodes, is_single_node_server)
if status is True:
self.cluster_util.wait_for_cb_collect_to_complete(rest)
self.cluster_util.copy_cb_collect_logs(rest, nodes, cluster,
log_path)
else:
self.log.error("API perform_cb_collect returned False")
def log_failure(self, message):
self.log.error(message)
self.summary.set_status("FAILED")
if self.test_failure is None:
self.test_failure = message
def validate_test_failure(self):
if self.test_failure is not None:
self.fail(self.test_failure)
def get_clusters(self):
return [self.cb_clusters[name] for name in self.cb_clusters.keys()]
def get_task(self):
return self.task
def get_task_mgr(self):
return self.task_manager
def init_sdk_pool_object(self):
self.sdk_client_pool = SDKClientPool()
DocLoaderUtils.sdk_client_pool = self.sdk_client_pool
def check_coredump_exist(self, servers, force_collect=False):
bin_cb = "/opt/couchbase/bin/"
lib_cb = "/opt/couchbase/var/lib/couchbase/"
# crash_dir = "/opt/couchbase/var/lib/couchbase/"
crash_dir_win = "c://CrashDumps"
result = False
self.data_sets = dict()
def find_index_of(str_list, sub_string):
for i in range(len(str_list)):
if sub_string in str_list[i]:
return i
return -1
def get_gdb(gdb_shell, dmp_path, dmp_name):
dmp_file = dmp_path + dmp_name
core_file = dmp_path + dmp_name.strip(".dmp") + ".core"
gdb_shell.execute_command("rm -rf " + core_file)
core_cmd = "/" + bin_cb + "minidump-2-core " + dmp_file + " > " + core_file
print("running: %s" % core_cmd)
gdb_shell.execute_command(core_cmd)
gdb = "gdb --batch {} -c {} -ex \"bt full\" -ex quit"\
.format(os.path.join(bin_cb, "memcached"), core_file)
print("running: %s" % gdb)
result = gdb_shell.execute_command(gdb)[0]
t_index = find_index_of(result, "Core was generated by")
result = result[t_index:]
result = " ".join(result)
return result
def get_full_thread_dump(shell):
cmd = 'gdb -p `(pidof memcached)` -ex "thread apply all bt" -ex detach -ex quit'
print("running: %s" % cmd)
thread_dump = shell.execute_command(cmd)[0]
index = find_index_of(
thread_dump, "Thread debugging using libthread_db enabled")
result = " ".join(thread_dump[index:])
return result
def run_cbanalyze_core(shell, core_file):
cbanalyze_core = os.path.join(bin_cb, "tools/cbanalyze-core")
cmd = '%s %s' % (cbanalyze_core, core_file)
print("running: %s" % cmd)
shell.execute_command(cmd)[0]
cbanalyze_log = core_file + ".log"
if shell.file_exists(os.path.dirname(cbanalyze_log),
os.path.basename(cbanalyze_log)):
log_path = TestInputSingleton.input.param(
"logs_folder", "/tmp")
shell.get_file(os.path.dirname(cbanalyze_log),
os.path.basename(cbanalyze_log), log_path)
def check_logs(grep_output_list):
"""
Check the grep's last line for the latest timestamp.
If this timestamp < start_timestamp of the test,
then return False (as the grep's output is from previous tests)
Note: This method works only if slave's time(timezone) matches
that of VM's. Else it won't be possible to compare timestamps
"""
last_line = grep_output_list[-1]
# eg: 2021-07-12T04:03:45
timestamp_regex = re.compile(
r"\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}")
match_obj = timestamp_regex.search(last_line)
if not match_obj:
self.log.critical("%s does not match any timestamp" %
last_line)
return True
timestamp = match_obj.group()
timestamp = datetime.strptime(timestamp, "%Y-%m-%dT%H:%M:%S")
self.log.info("Comparing timestamps: Log's latest timestamp: %s, "
"Test's start timestamp is %s" %
(timestamp, self.start_timestamp))
if timestamp > self.start_timestamp:
return True
else:
return False
for idx, server in enumerate(servers):
shell = RemoteMachineShellConnection(server)
shell.extract_remote_info()
crash_dir = lib_cb + "crash/"
if shell.info.type.lower() == "windows":
crash_dir = crash_dir_win
if int(server.port) in range(ClusterRun.port,
ClusterRun.port + 10):
crash_dir = os.path.join(
TestInputSingleton.input.servers[0].cli_path, "ns_server",
"data", "n_%s" % str(idx), "crash")
dmp_files = shell.execute_command("ls -lt " + crash_dir)[0]
dmp_files = [f for f in dmp_files if ".core" not in f]
dmp_files = [f for f in dmp_files if "total" not in f]
dmp_files = [f.split()[-1] for f in dmp_files if ".core" not in f]
dmp_files = [f.strip("\n") for f in dmp_files]
if dmp_files:
print("#" * 30)
print("%s: %d core dump seen" % (server.ip, len(dmp_files)))
print("%s: Stack Trace of first crash - %s\n%s" %
(server.ip, dmp_files[-1],
get_gdb(shell, crash_dir, dmp_files[-1])))
print("#" * 30)
result = get_full_thread_dump(shell)
print(result)
# print("#"*30)
# run_cbanalyze_core(shell, crash_dir + dmp_files[-1].strip(".dmp") + ".core")
# print("#"*30)
if self.stop_server_on_crash:
shell.stop_couchbase()
result = True
else:
self.log.debug(server.ip + ": No crash files found")
logs_dir = lib_cb + "logs/"
if int(server.port) in range(ClusterRun.port,
ClusterRun.port + 10):
logs_dir = os.path.join(
TestInputSingleton.input.servers[0].cli_path, "ns_server",
"logs", "n_%s" % str(idx))
# Perform log file searching based on the input yaml config
yaml = YAML()
with open("lib/couchbase_helper/error_log_config.yaml", "r") as fp:
y_data = yaml.load(fp.read())
for file_data in y_data["file_name_patterns"]:
log_files = shell.execute_command(
"ls " + os.path.join(logs_dir, file_data['file']))[0]
if len(log_files) == 0:
self.log.debug("%s: No '%s' files found" %
(server.ip, file_data['file']))
continue
if 'target_file_index' in file_data:
log_files = [
log_files[int(file_data['target_file_index'])]
]
for log_file in log_files:
log_file = log_file.strip("\n")
for grep_pattern in file_data['grep_for']:
grep_for_str = grep_pattern['string']
err_pattern = exclude_pattern = None
if 'error_patterns' in grep_pattern:
err_pattern = grep_pattern['error_patterns']
if 'exclude_patterns' in grep_pattern:
exclude_pattern = grep_pattern['exclude_patterns']
cmd_to_run = "grep -r '%s' %s" \
% (grep_for_str, log_file)
if exclude_pattern is not None:
for pattern in exclude_pattern:
cmd_to_run += " | grep -v '%s'" % pattern
grep_output = shell.execute_command(cmd_to_run)[0]
if grep_output and check_logs(grep_output):
regex = r"(\bkvstore-\d+)"
grep_str = "".join(grep_output)
kvstores = list(set(re.findall(regex, grep_str)))
self.data_sets[server] = kvstores
grep_str = None
if err_pattern is not None:
for pattern in err_pattern:
index = find_index_of(grep_output, pattern)
grep_output = grep_output[:index]
if grep_output:
self.log.info("unwanted messages in %s" %
log_file)
if check_logs(grep_output):
self.log.critical(
"%s: Found '%s' logs - %s" %
(server.ip, grep_for_str,
"".join(grep_output)))
result = True
break
else:
if grep_output \
and check_logs(grep_output):
self.log.info("unwanted messages in %s" %
log_file)
self.log.critical(
"%s: Found '%s' logs - %s" %
(server.ip, grep_for_str, grep_output))
result = True
break
if result is True:
if self.stop_server_on_crash:
shell.stop_couchbase()
break
shell.disconnect()
if result and force_collect and not self.stop_server_on_crash:
self.fetch_cb_collect_logs()
self.get_cbcollect_info = False
if (self.is_test_failed() or result) and self.collect_data:
self.copy_data_on_slave()
return result
def copy_data_on_slave(self, servers=None):
log_path = TestInputSingleton.input.param("logs_folder", "/tmp")
if servers is None:
servers = self.cluster.nodes_in_cluster
for node in servers:
if "kv" not in node.services.lower():
servers.remove(node)
if type(servers) is not list:
servers = [servers]
remote_path = RestConnection(servers[0]).get_data_path()
file_path = os.path.join(remote_path, self.cluster.buckets[0].name)
file_name = self.cluster.buckets[0].name + ".tar.gz"
def get_tar(remotepath, filepath, filename, servers, todir="."):
if type(servers) is not list:
servers = [servers]
for server in servers:
shell = RemoteMachineShellConnection(server)
_ = shell.execute_command("tar -zcvf %s.tar.gz %s" %
(filepath, filepath))
file_check = shell.file_exists(remotepath, filename)
if not file_check:
self.log.error(
"Tar File {} doesn't exist".format(filename))
tar_file_copied = shell.get_file(remotepath, filename, todir)
if not tar_file_copied:
self.log.error("Failed to copy Tar file")
_ = shell.execute_command("rm -rf %s.tar.gz" % filepath)
copy_path_msg_format = "Copying data, Server :: %s, Path :: %s"
'''
Temporarily enabling data copy
of all nodes irrespective of nodes in
data_sets
'''
if False and self.data_sets and self.bucket_storage == "magma":
self.log.critical("data_sets ==> {}".format(self.data_sets))
wal_tar = "wal.tar.gz"
config_json_tar = "config.json.tar.gz"
for server, kvstores in self.data_sets.items():
shell = RemoteMachineShellConnection(server)
if not kvstores:
copy_to_path = os.path.join(log_path,
server.ip.replace(".", "_"))
if not os.path.isdir(copy_to_path):
os.makedirs(copy_to_path, 0o777)
self.log.info(copy_path_msg_format %
(server.ip, copy_to_path))
get_tar(remote_path,
file_path,
file_name,
server,
todir=copy_to_path)
else:
for kvstore in kvstores:
if int(kvstore.split("-")[1]) >= self.vbuckets:
continue
kvstore_path = shell.execute_command(
"find %s -type d -name '%s'" %
(remote_path, kvstore))[0][0]
magma_dir = kvstore_path.split(kvstore)[0]
wal_path = kvstore_path.split(kvstore)[0] + "wal"
config_json_path = kvstore_path.split(kvstore)[0]
+ "config.json"
kvstore_path = kvstore_path.split(kvstore)[0] + kvstore
kvstore_tar = kvstore + ".tar.gz"
copy_to_path = os.path.join(log_path, kvstore)
if not os.path.isdir(copy_to_path):
os.makedirs(copy_to_path, 0o777)
self.log.info(copy_path_msg_format %
(server.ip, copy_to_path))
get_tar(magma_dir,
kvstore_path,
kvstore_tar,
server,
todir=copy_to_path)
get_tar(magma_dir,
wal_path,
wal_tar,
server,
todir=copy_to_path)
get_tar(magma_dir,
config_json_path,
config_json_tar,
server,
todir=copy_to_path)
else:
for server in servers:
copy_to_path = os.path.join(log_path,
server.ip.replace(".", "_"))
if not os.path.isdir(copy_to_path):
os.makedirs(copy_to_path, 0o777)
self.log.info(copy_path_msg_format % (server.ip, copy_to_path))
get_tar(remote_path,
file_path,
file_name,
server,
todir=copy_to_path)