TrueCloudLab/frostfs-testlib
25
0
Fork 20
Code Issues Pull requests 2 Releases 1 Activity Actions

Added wait_for_node_to_be_ready to starting stopped storage nodes #29

Merged
abereziny merged 1 commit from ylukoyan/frostfs-testlib:fixture_start_storage into master 2023-05-29 14:01:58 +00:00
Conversation 0 Commits 1 Files changed 1 +2
1 changed files with 2 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
src/frostfs_testlib/storage/controllers/cluster_state_controller.py
View file

@ -7,6 +7,7 @@ from frostfs_testlib.reporter import get_reporter
from frostfs_testlib.shell import CommandOptions, Shell from frostfs_testlib.shell import CommandOptions, Shell
from frostfs_testlib.steps import epoch from frostfs_testlib.steps import epoch
from frostfs_testlib.storage.cluster import Cluster, ClusterNode, StorageNode from frostfs_testlib.storage.cluster import Cluster, ClusterNode, StorageNode
from frostfs_testlib.steps.node_management import wait_for_node_to_be_ready
from frostfs_testlib.storage.controllers.disk_controller import DiskController from frostfs_testlib.storage.controllers.disk_controller import DiskController
from frostfs_testlib.testing.test_control import run_optionally, wait_for_success from frostfs_testlib.testing.test_control import run_optionally, wait_for_success
from frostfs_testlib.utils.failover_utils import ( from frostfs_testlib.utils.failover_utils import (
@ -90,6 +91,7 @@ class ClusterStateController:
def start_stopped_storage_services(self): def start_stopped_storage_services(self):
for node in self.stopped_storage_nodes: for node in self.stopped_storage_nodes:
node.start_service() node.start_service()
wait_all_storage_nodes_returned(self.shell, self.cluster)
abereziny marked this conversation as resolved Outdated
abereziny commented 2023-05-26 08:26:36 +00:00
Outdated
Review
Copy link

This is potentially bad change.
Let say you have 4 services turned off.

You start first one and wait for READY status.
Since other 3 services are still down - this one will also never be ready.

If you want to wait for it to be ready, you should use
wait_all_storage_nodes_returned(self.shell, self.cluster)
after all services were started.

This is potentially bad change. Let say you have 4 services turned off. You start first one and wait for READY status. Since other 3 services are still down - this one will also never be ready. If you want to wait for it to be ready, you should use `wait_all_storage_nodes_returned(self.shell, self.cluster)` after all services were started.
👀 1
self.stopped_storage_nodes = [] self.stopped_storage_nodes = []
@run_optionally(optionals.OPTIONAL_FAILOVER_ENABLED) @run_optionally(optionals.OPTIONAL_FAILOVER_ENABLED)