Skip to content

Fix: reset shares stuck in extending_error to available #331

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
skook1 wants to merge 1 commit into
base: stable/2025.1-m3
Choose a base branch
from
Open

Fix: reset shares stuck in extending_error to available #331

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
118 changes: 110 additions & 8 deletions manila/share/manager.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1760,11 +1760,26 @@ def migration_driver_continue(self, context):

for instance in instances:

share = self.db.share_get(context, instance['share_id'])

if instance['status'] != constants.STATUS_MIGRATING:
if share['task_state'] == (
constants.TASK_STATE_MIGRATION_DRIVER_IN_PROGRESS):
LOG.warning(
"Share instance %(instance_id)s (share %(share_id)s) "
"is skipped by migration_driver_continue because its "
"status is %(status)s (expected %(expected)s), while "
"task_state is still %(task_state)s. The migration "
"task_state may be orphaned - this typically happens "
"when extend_share/shrink_share overwrote the "
"migrating status.",
{'instance_id': instance['id'],
'share_id': instance['share_id'],
'status': instance['status'],
'expected': constants.STATUS_MIGRATING,
'task_state': share['task_state']})
continue

share = self.db.share_get(context, instance['share_id'])

if share['task_state'] == (
constants.TASK_STATE_MIGRATION_DRIVER_IN_PROGRESS):

Expand Down Expand Up @@ -5342,6 +5357,42 @@ def _delete_encryption_keys_quota(self, context):
context, reservations, project_id=context.project_id,
)

def _detect_migration_status(self, share, share_server):
"""Detect whether the share is being migrated."""

share_task_state = share.get('task_state')
was_share_migrating = share_task_state in constants.BUSY_TASK_STATES

was_server_migrating = False
if share_server is not None and hasattr(share_server, 'get'):
server_status = share_server.get('status')
server_task_state = share_server.get('task_state')
was_server_migrating = (
server_status == constants.STATUS_SERVER_MIGRATING
or server_task_state in constants.BUSY_TASK_STATES)

if was_server_migrating:
return True, constants.STATUS_SERVER_MIGRATING
if was_share_migrating:
return True, constants.STATUS_MIGRATING
return False, share['status']

def _migration_still_running(self, context, share, share_server):
"""Re-check migration state right before a final status write."""

fresh_share = self.db.share_get(context, share['id'])
fresh_server = None
if share_server:
try:
fresh_server = self.db.share_server_get(
context, share_server['id'])
except exception.ShareServerNotFound:
fresh_server = None
_, current_original = self._detect_migration_status(
fresh_share, fresh_server)
return current_original in (constants.STATUS_MIGRATING,
constants.STATUS_SERVER_MIGRATING)

@run_concurrently
@add_hooks
@utils.require_driver_initialized
Expand All @@ -5353,6 +5404,9 @@ def extend_share(self, context, share_id, new_size, reservations):
project_id = share['project_id']
user_id = share['user_id']

was_migrating, original_status = self._detect_migration_status(
share, share_server)

self._notify_about_share_usage(context, share,
share_instance, "extend.start")

Expand All @@ -5369,9 +5423,15 @@ def extend_share(self, context, share_id, new_size, reservations):
resource_id=share_id,
detail=message_field.Detail.DRIVER_FAILED_EXTEND)
try:
still_migrating = (
was_migrating
and self._migration_still_running(
context, share, share_server))
error_status = (original_status if still_migrating
else constants.STATUS_EXTENDING_ERROR)
self.db.share_update(
context, share['id'],
{'status': constants.STATUS_EXTENDING_ERROR}
{'status': error_status}
)
raise exception.ShareExtendingError(
reason=str(e), share_id=share_id)
Expand All @@ -5390,11 +5450,16 @@ def extend_share(self, context, share_id, new_size, reservations):
user_id=user_id, share_type_id=share_instance['share_type_id'],
)

still_migrating = (
was_migrating
and self._migration_still_running(context, share, share_server))
success_status = (original_status if still_migrating
else constants.STATUS_AVAILABLE.lower())
share_update = {
'size': int(new_size),
# NOTE(u_glide): translation to lower case should be removed in
# a row with usage of upper case of share statuses in all places
'status': constants.STATUS_AVAILABLE.lower()
'status': success_status
}
share = self.db.share_update(context, share['id'], share_update)

Expand All @@ -5417,6 +5482,9 @@ def shrink_share(self, context, share_id, new_size):
context, share['id'])
supports_replication = len(replicas) > 0

was_migrating, original_status = self._detect_migration_status(
share, share_server)

self._notify_about_share_usage(context, share,
share_instance, "shrink.start")

Expand All @@ -5432,7 +5500,13 @@ def error_occurred(exc, msg, status=constants.STATUS_SHRINKING_ERROR):
resource_id=share['id'],
detail=message_field.Detail.DRIVER_FAILED_SHRINK)
LOG.warning(msg, resource=share)
self.db.share_update(context, share['id'], {'status': status})
still_migrating = (
was_migrating
and self._migration_still_running(
context, share, share_server))
final_status = original_status if still_migrating else status
self.db.share_update(
context, share['id'], {'status': final_status})

raise exception.ShareShrinkingError(
reason=str(exc), share_id=share_id)
Expand Down Expand Up @@ -5497,9 +5571,14 @@ def error_occurred(exc, msg, status=constants.STATUS_SHRINKING_ERROR):
user_id=user_id, share_type_id=share_instance['share_type_id'],
)

still_migrating = (
was_migrating
and self._migration_still_running(context, share, share_server))
success_status = (original_status if still_migrating
else constants.STATUS_AVAILABLE)
share_update = {
'size': new_size,
'status': constants.STATUS_AVAILABLE
'status': success_status
}
share = self.db.share_update(context, share['id'], share_update)

Expand Down Expand Up @@ -6352,8 +6431,31 @@ def _update_resource_status(self, context, status, task_state=None,
if task_state:
fields['task_state'] = task_state
if share_instance_ids:
self.db.share_instance_status_update(
context, share_instance_ids, fields)

preserved_statuses = (
constants.STATUS_EXTENDING_ERROR,
constants.STATUS_SHRINKING_ERROR,
)
ids_to_update = []
for instance_id in share_instance_ids:
instance = self.db.share_instance_get(context, instance_id)
if instance['status'] in preserved_statuses:
LOG.warning(
"Not overwriting share instance %(id)s status "
"%(status)s during resource status update to "
"%(new_status)s - the error state is preserved for "
"operator visibility.",
{'id': instance_id,
'status': instance['status'],
'new_status': status})
if task_state:
self.db.share_instance_update(
context, instance_id, {'task_state': task_state})
else:
ids_to_update.append(instance_id)
if ids_to_update:
self.db.share_instance_status_update(
context, ids_to_update, fields)
if snapshot_instance_ids:
self.db.share_snapshot_instances_status_update(
context, snapshot_instance_ids, fields)
Expand Down
Loading