frostfs-testcases/pytest_tests/testsuites/shard/test_control_shard.py
2024-11-26 19:42:53 +03:00

149 lines
6.2 KiB
Python

import json
import os
import allure
import pytest
from frostfs_testlib import reporter
from frostfs_testlib.cli import FrostfsCli
from frostfs_testlib.resources.cli import CLI_DEFAULT_TIMEOUT
from frostfs_testlib.shell.interfaces import Shell
from frostfs_testlib.steps.cli.object import get_object, get_object_nodes, put_object
from frostfs_testlib.storage.cluster import Cluster, ClusterNode, StorageNode
from frostfs_testlib.storage.controllers import ClusterStateController, ShardsWatcher
from frostfs_testlib.storage.controllers.state_managers.config_state_manager import ConfigStateManager
from frostfs_testlib.storage.dataclasses.shard import Shard
from frostfs_testlib.storage.dataclasses.wallet import WalletInfo
from frostfs_testlib.testing import parallel, wait_for_success
from frostfs_testlib.testing.cluster_test_base import ClusterTestBase
from frostfs_testlib.utils.file_utils import generate_file
from ...helpers.container_request import PUBLIC_WITH_POLICY, requires_container
def set_shard_rw_mode(node: ClusterNode):
watcher = ShardsWatcher(node)
shards = watcher.get_shards()
for shard in shards:
watcher.set_shard_mode(shard["shard_id"], mode="read-write")
watcher.await_for_all_shards_status(status="read-write")
@pytest.fixture()
@allure.title("Revert all shards mode")
def revert_all_shards_mode(cluster: Cluster) -> None:
yield
parallel(set_shard_rw_mode, cluster.cluster_nodes)
@pytest.fixture()
def object_id(client_shell: Shell, cluster: Cluster, container: str, default_wallet: WalletInfo, max_object_size: int) -> str:
with reporter.step("Create container, and put object"):
file = generate_file(round(max_object_size * 0.8))
oid = put_object(default_wallet, file, container, client_shell, cluster.default_rpc_endpoint)
return oid
@pytest.fixture()
def node_with_object(cluster: Cluster, container: str, object_id: str) -> ClusterNode:
with reporter.step("Search node with object"):
nodes = get_object_nodes(cluster, container, object_id, cluster.cluster_nodes[0])
return nodes[0]
@pytest.fixture()
@wait_for_success(180, 30, title="Search object in system")
def object_path_on_node(object_id: str, container: str, node_with_object: ClusterNode) -> str:
oid_path = f"{object_id[0]}/{object_id[1]}/{object_id[2]}/{object_id[3]}"
object_path = None
with reporter.step("Search object file"):
node_shell = node_with_object.storage_node.host.get_shell()
data_path = node_with_object.storage_node.get_data_directory()
all_datas = node_shell.exec(f"ls -la {data_path}/data | awk '{{ print $9 }}'").stdout.strip()
for data_dir in all_datas.replace(".", "").strip().split("\n"):
check_dir = node_shell.exec(f" [ -d {data_path}/data/{data_dir}/data/{oid_path} ] && echo 1 || echo 0").stdout
if "1" in check_dir:
object_path = f"{data_path}/data/{data_dir}/data/{oid_path}"
object_name = f"{object_id[4:]}.{container}"
break
assert object_path is not None, f"{object_id} object not found in directory - {data_path}/data"
return os.path.join(object_path, object_name)
@pytest.fixture()
def erroneous_object_id(object_id: str, object_path_on_node: str, node_with_object: ClusterNode):
with reporter.step("Block read file"):
node_with_object.host.get_shell().exec(f"chmod a-r {object_path_on_node}")
yield object_id
with reporter.step("Restore file access"):
node_with_object.host.get_shell().exec(f"chmod +r {object_path_on_node}")
@pytest.fixture()
def change_config_storage(cluster_state_controller: ClusterStateController):
with reporter.step("Change threshold error shards"):
cluster_state_controller.manager(ConfigStateManager).set_on_all_nodes(
service_type=StorageNode, values={"storage:shard_ro_error_threshold": "5"}
)
yield
with reporter.step("Restore threshold error shards"):
cluster_state_controller.manager(ConfigStateManager).revert_all()
@pytest.mark.nightly
@pytest.mark.shard
class TestControlShard(ClusterTestBase):
@staticmethod
def get_shards_from_cli(node: StorageNode) -> list[Shard]:
wallet_path = node.get_remote_wallet_path()
wallet_password = node.get_wallet_password()
control_endpoint = node.get_control_endpoint()
cli_config = node.host.get_cli_config("frostfs-cli")
cli = FrostfsCli(node.host.get_shell(), cli_config.exec_path)
result = cli.shards.list(
endpoint=control_endpoint,
wallet=wallet_path,
wallet_password=wallet_password,
json_mode=True,
timeout=CLI_DEFAULT_TIMEOUT,
)
return [Shard.from_object(shard) for shard in json.loads(result.stdout.split(">", 1)[1])]
@allure.title("All shards are available")
def test_control_shard(self, cluster: Cluster):
for storage_node in cluster.storage_nodes:
shards_from_config = storage_node.get_shards()
shards_from_cli = self.get_shards_from_cli(storage_node)
assert set(shards_from_config) == set(shards_from_cli)
@allure.title("Shard become read-only when errors exceeds threshold")
@pytest.mark.failover
@requires_container(PUBLIC_WITH_POLICY("REP 1 CBF 1", short_name="REP 1"))
def test_shard_errors(
self,
default_wallet: WalletInfo,
container: str,
node_with_object: ClusterNode,
erroneous_object_id: str,
object_path_on_node: str,
change_config_storage: None,
revert_all_shards_mode: None,
):
with reporter.step("Get object, expect 6 errors"):
for _ in range(6):
with pytest.raises(RuntimeError):
get_object(default_wallet, container, erroneous_object_id, self.shell, node_with_object.storage_node.get_rpc_endpoint())
with reporter.step("Check shard status"):
for shard in ShardsWatcher(node_with_object).get_shards():
if shard["blobstor"][1]["path"] in object_path_on_node:
with reporter.step(f"Shard {shard['shard_id']} should be in read-only mode"):
assert shard["mode"] == "read-only"
break