forked from TrueCloudLab/frostfs-testcases
293 lines
13 KiB
Python
293 lines
13 KiB
Python
import random
|
|
import re
|
|
|
|
import allure
|
|
import pytest
|
|
from frostfs_testlib import reporter
|
|
from frostfs_testlib.steps.cli.container import create_container, delete_container, search_nodes_with_container
|
|
from frostfs_testlib.steps.cli.object import delete_object, lock_object, put_object, put_object_to_random_node
|
|
from frostfs_testlib.steps.metrics import check_metrics_counter, get_metrics_value
|
|
from frostfs_testlib.steps.storage_policy import get_nodes_with_object
|
|
from frostfs_testlib.storage.cluster import Cluster, ClusterNode
|
|
from frostfs_testlib.storage.controllers.cluster_state_controller import ClusterStateController
|
|
from frostfs_testlib.storage.dataclasses.object_size import ObjectSize
|
|
from frostfs_testlib.storage.dataclasses.wallet import WalletInfo
|
|
from frostfs_testlib.testing.cluster_test_base import ClusterTestBase
|
|
from frostfs_testlib.utils.file_utils import generate_file
|
|
|
|
|
|
class TestObjectMetrics(ClusterTestBase):
|
|
@allure.title("Object metrics of removed container (obj_size={object_size})")
|
|
def test_object_metrics_removed_container(self, object_size: ObjectSize, default_wallet: WalletInfo, cluster: Cluster):
|
|
file_path = generate_file(object_size.value)
|
|
placement_policy = "REP 2 IN X CBF 2 SELECT 2 FROM * AS X"
|
|
copies = 2
|
|
|
|
with reporter.step(f"Create container with policy {placement_policy}"):
|
|
cid = create_container(default_wallet, self.shell, cluster.default_rpc_endpoint, placement_policy)
|
|
|
|
with reporter.step("Put object to random node"):
|
|
oid = put_object_to_random_node(default_wallet, file_path, cid, self.shell, cluster)
|
|
|
|
with reporter.step("Check metric appears in node where the object is located"):
|
|
object_storage_nodes = get_nodes_with_object(cid, oid, self.shell, cluster.storage_nodes)
|
|
object_nodes = [cluster_node for cluster_node in cluster.cluster_nodes if cluster_node.storage_node in object_storage_nodes]
|
|
|
|
check_metrics_counter(
|
|
object_nodes,
|
|
counter_exp=copies,
|
|
command="frostfs_node_engine_container_objects_total",
|
|
cid=cid,
|
|
type="user",
|
|
)
|
|
|
|
with reporter.step("Delete container"):
|
|
delete_container(default_wallet, cid, shell=self.shell, endpoint=self.cluster.default_rpc_endpoint)
|
|
|
|
with reporter.step("Tick Epoch"):
|
|
self.tick_epochs(epochs_to_tick=2, wait_block=2)
|
|
|
|
with reporter.step("Check metrics of removed containers doesn't appear in the storage node"):
|
|
check_metrics_counter(object_nodes, counter_exp=0, command="frostfs_node_engine_container_objects_total", cid=cid, type="user")
|
|
check_metrics_counter(object_nodes, counter_exp=0, command="frostfs_node_engine_container_size_byte", cid=cid)
|
|
|
|
for node in object_nodes:
|
|
all_metrics = node.metrics.storage.get_metrics_search_by_greps(command="frostfs_node_engine_container_size_byte")
|
|
assert cid not in all_metrics.stdout, "metrics of removed containers shouldn't appear in the storage node"
|
|
|
|
@allure.title("Object metrics, locked object (obj_size={object_size}, policy={placement_policy})")
|
|
@pytest.mark.parametrize("placement_policy", ["REP 1 IN X CBF 1 SELECT 1 FROM * AS X", "REP 2 IN X CBF 2 SELECT 2 FROM * AS X"])
|
|
def test_object_metrics_blocked_object(
|
|
self, object_size: ObjectSize, default_wallet: WalletInfo, cluster: Cluster, placement_policy: str
|
|
):
|
|
file_path = generate_file(object_size.value)
|
|
metric_step = int(re.search(r"REP\s(\d+)", placement_policy).group(1))
|
|
|
|
with reporter.step(f"Create container with policy {placement_policy}"):
|
|
cid = create_container(default_wallet, self.shell, cluster.default_rpc_endpoint, placement_policy)
|
|
|
|
with reporter.step("Search container nodes"):
|
|
container_nodes = search_nodes_with_container(
|
|
wallet=default_wallet,
|
|
cid=cid,
|
|
shell=self.shell,
|
|
endpoint=self.cluster.default_rpc_endpoint,
|
|
cluster=cluster,
|
|
)
|
|
|
|
with reporter.step("Get current metrics for metric_type=user"):
|
|
objects_metric_counter = 0
|
|
for node in container_nodes:
|
|
objects_metric_counter += get_metrics_value(node, command="frostfs_node_engine_objects_total", type="user")
|
|
|
|
with reporter.step("Put object to container node"):
|
|
oid = put_object(default_wallet, file_path, cid, self.shell, container_nodes[0].storage_node.get_rpc_endpoint())
|
|
|
|
with reporter.step(f"Check metric user 'the counter should increase by {metric_step}'"):
|
|
objects_metric_counter += metric_step
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=objects_metric_counter,
|
|
command="frostfs_node_engine_objects_total",
|
|
type="user",
|
|
)
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=metric_step,
|
|
command="frostfs_node_engine_container_objects_total",
|
|
cid=cid,
|
|
type="user",
|
|
)
|
|
|
|
with reporter.step("Delete object"):
|
|
delete_object(default_wallet, cid, oid, self.shell, self.cluster.default_rpc_endpoint)
|
|
|
|
with reporter.step(f"Check metric user 'the counter should decrease by {metric_step}'"):
|
|
objects_metric_counter -= metric_step
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=objects_metric_counter,
|
|
command="frostfs_node_engine_objects_total",
|
|
type="user",
|
|
)
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=0,
|
|
command="frostfs_node_engine_container_objects_total",
|
|
cid=cid,
|
|
type="user",
|
|
)
|
|
|
|
with reporter.step("Put object and lock it to next epoch"):
|
|
oid = put_object(default_wallet, file_path, cid, self.shell, container_nodes[0].storage_node.get_rpc_endpoint())
|
|
current_epoch = self.get_epoch()
|
|
lock_object(
|
|
default_wallet,
|
|
cid,
|
|
oid,
|
|
self.shell,
|
|
container_nodes[0].storage_node.get_rpc_endpoint(),
|
|
expire_at=current_epoch + 1,
|
|
)
|
|
|
|
with reporter.step(f"Check metric user 'the counter should increase by {metric_step}'"):
|
|
objects_metric_counter += metric_step
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=objects_metric_counter,
|
|
command="frostfs_node_engine_objects_total",
|
|
type="user",
|
|
)
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=metric_step,
|
|
command="frostfs_node_engine_container_objects_total",
|
|
cid=cid,
|
|
type="user",
|
|
)
|
|
|
|
with reporter.step(f"Wait until remove locking 'the counter doesn't change'"):
|
|
self.tick_epochs(epochs_to_tick=2)
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=objects_metric_counter,
|
|
command="frostfs_node_engine_objects_total",
|
|
type="user",
|
|
)
|
|
|
|
with reporter.step("Delete object"):
|
|
delete_object(default_wallet, cid, oid, self.shell, self.cluster.default_rpc_endpoint)
|
|
|
|
with reporter.step(f"Check metric user 'the counter should decrease by {metric_step}'"):
|
|
objects_metric_counter -= metric_step
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=objects_metric_counter,
|
|
command="frostfs_node_engine_objects_total",
|
|
type="user",
|
|
)
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=0,
|
|
command="frostfs_node_engine_container_objects_total",
|
|
cid=cid,
|
|
type="user",
|
|
)
|
|
|
|
with reporter.step("Put object with expire_at"):
|
|
current_epoch = self.get_epoch()
|
|
oid = put_object(
|
|
default_wallet,
|
|
file_path,
|
|
cid,
|
|
self.shell,
|
|
container_nodes[0].storage_node.get_rpc_endpoint(),
|
|
expire_at=current_epoch + 1,
|
|
)
|
|
|
|
with reporter.step(f"Check metric user 'the counter should increase by {metric_step}'"):
|
|
objects_metric_counter += metric_step
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=objects_metric_counter,
|
|
command="frostfs_node_engine_objects_total",
|
|
type="user",
|
|
)
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=metric_step,
|
|
command="frostfs_node_engine_container_objects_total",
|
|
cid=cid,
|
|
type="user",
|
|
)
|
|
|
|
with reporter.step("Tick Epoch"):
|
|
self.tick_epochs(epochs_to_tick=2)
|
|
|
|
with reporter.step(f"Check metric user 'the counter should decrease by {metric_step}'"):
|
|
objects_metric_counter -= metric_step
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=objects_metric_counter,
|
|
command="frostfs_node_engine_objects_total",
|
|
type="user",
|
|
)
|
|
check_metrics_counter(
|
|
container_nodes,
|
|
counter_exp=0,
|
|
command="frostfs_node_engine_container_objects_total",
|
|
cid=cid,
|
|
type="user",
|
|
)
|
|
|
|
@allure.title("Object metrics, stop the node (obj_size={object_size})")
|
|
def test_object_metrics_stop_node(
|
|
self,
|
|
object_size: ObjectSize,
|
|
default_wallet: WalletInfo,
|
|
cluster_state_controller: ClusterStateController,
|
|
):
|
|
placement_policy = "REP 2 IN X CBF 2 SELECT 2 FROM * AS X"
|
|
file_path = generate_file(object_size.value)
|
|
copies = 2
|
|
|
|
with reporter.step(f"Create container with policy {placement_policy}"):
|
|
cid = create_container(default_wallet, self.shell, self.cluster.default_rpc_endpoint, placement_policy)
|
|
|
|
with reporter.step(f"Check object metrics in container 'should be zero'"):
|
|
check_metrics_counter(
|
|
self.cluster.cluster_nodes,
|
|
counter_exp=0,
|
|
command="frostfs_node_engine_container_objects_total",
|
|
type="user",
|
|
cid=cid,
|
|
)
|
|
|
|
with reporter.step("Get current metrics for each nodes"):
|
|
objects_metric_counter: dict[ClusterNode:int] = {}
|
|
for node in self.cluster.cluster_nodes:
|
|
objects_metric_counter[node] = get_metrics_value(node, command="frostfs_node_engine_objects_total", type="user")
|
|
|
|
with reporter.step("Put object"):
|
|
oid = put_object(default_wallet, file_path, cid, self.shell, self.cluster.default_rpc_endpoint)
|
|
|
|
with reporter.step("Get object nodes"):
|
|
object_storage_nodes = get_nodes_with_object(cid, oid, self.shell, self.cluster.storage_nodes)
|
|
object_nodes = [
|
|
cluster_node for cluster_node in self.cluster.cluster_nodes if cluster_node.storage_node in object_storage_nodes
|
|
]
|
|
|
|
with reporter.step(f"Check metrics in object nodes 'the counter should increase by {copies}'"):
|
|
counter_exp = sum(objects_metric_counter[node] for node in object_nodes) + copies
|
|
check_metrics_counter(object_nodes, counter_exp=counter_exp, command="frostfs_node_engine_objects_total", type="user")
|
|
check_metrics_counter(
|
|
object_nodes,
|
|
counter_exp=copies,
|
|
command="frostfs_node_engine_container_objects_total",
|
|
type="user",
|
|
cid=cid,
|
|
)
|
|
|
|
with reporter.step(f"Select node to stop"):
|
|
node_to_stop = random.choice(object_nodes)
|
|
alive_nodes = set(object_nodes).difference({node_to_stop})
|
|
|
|
with reporter.step(f"Stop the node, wait until the object is replicated to another node"):
|
|
cluster_state_controller.stop_node_host(node_to_stop, "hard")
|
|
objects_metric_counter[node_to_stop] += 1
|
|
|
|
with reporter.step(f"Check metric in alive nodes 'the counter should increase'"):
|
|
counter_exp = sum(objects_metric_counter[node] for node in alive_nodes)
|
|
check_metrics_counter(alive_nodes, ">=", counter_exp, command="frostfs_node_engine_objects_total", type="user")
|
|
|
|
with reporter.step("Start node"):
|
|
cluster_state_controller.start_node_host(node_to_stop)
|
|
|
|
with reporter.step(f"Check metric in restarted node, 'the counter doesn't change'"):
|
|
check_metrics_counter(
|
|
object_nodes,
|
|
counter_exp=copies,
|
|
command="frostfs_node_engine_container_objects_total",
|
|
type="user",
|
|
cid=cid,
|
|
)
|