async def list_placement_groups(self, *, option: ListApiOptions) -> ListApiResponse:
"""List all placement group information from the cluster.
Returns:
{pg_id -> pg_data_in_dict}
pg_data_in_dict's schema is in PlacementGroupState
"""
try:
reply = await self._client.get_all_placement_group_info(
timeout=option.timeout
)
except DataSourceUnavailable:
raise DataSourceUnavailable(GCS_QUERY_FAILURE_WARNING)
result = []
for message in reply.placement_group_table_data:
data = self._message_to_dict(
message=message,
fields_to_decode=["placement_group_id"],
)
result.append(data)
result = self._filter(result, option.filters, PlacementGroupState)
# Sort to make the output deterministic.
result.sort(key=lambda entry: entry["placement_group_id"])
return ListApiResponse(result=list(islice(result, option.limit)))
async def list_workers(self, *, option: ListApiOptions) -> ListApiResponse:
"""List all worker information from the cluster.
Returns:
{worker_id -> worker_data_in_dict}
worker_data_in_dict's schema is in WorkerState
"""
try:
reply = await self._client.get_all_worker_info(
timeout=option.timeout)
except DataSourceUnavailable:
raise DataSourceUnavailable(GCS_QUERY_FAILURE_WARNING)
result = []
for message in reply.worker_table_data:
data = self._message_to_dict(
message=message, fields_to_decode=["worker_id", "raylet_id"])
data["worker_id"] = data["worker_address"]["worker_id"]
data["node_id"] = data["worker_address"]["raylet_id"]
data["ip"] = data["worker_address"]["ip_address"]
result.append(data)
result = self._filter(result, option.filters, WorkerState,
option.detail)
# Sort to make the output deterministic.
result.sort(key=lambda entry: entry["worker_id"])
result = list(islice(result, option.limit))
return ListApiResponse(
result=result,
total=reply.total,
)
async def list_nodes(self, *, option: ListApiOptions) -> ListApiResponse:
"""List all node information from the cluster.
Returns:
{node_id -> node_data_in_dict}
node_data_in_dict's schema is in NodeState
"""
try:
reply = await self._client.get_all_node_info(timeout=option.timeout
)
except DataSourceUnavailable:
raise DataSourceUnavailable(GCS_QUERY_FAILURE_WARNING)
result = []
for message in reply.node_info_list:
data = self._message_to_dict(message=message,
fields_to_decode=["node_id"])
data["node_ip"] = data["node_manager_address"]
result.append(data)
result = self._filter(result, option.filters, NodeState, option.detail)
# Sort to make the output deterministic.
result.sort(key=lambda entry: entry["node_id"])
total_nodes = len(result)
result = list(islice(result, option.limit))
return ListApiResponse(
result=result,
# No reason to truncate node because they are usually small.
total=total_nodes,
)
def list_jobs(self, *, option: ListApiOptions) -> ListApiResponse:
# TODO(sang): Support limit & timeout & async calls.
try:
result = self._client.get_job_info()
except DataSourceUnavailable:
raise DataSourceUnavailable(GCS_QUERY_FAILURE_WARNING)
return ListApiResponse(result=result)
async def list_nodes(self, *, option: ListApiOptions) -> ListApiResponse:
"""List all node information from the cluster.
Returns:
{node_id -> node_data_in_dict}
node_data_in_dict's schema is in NodeState
"""
try:
reply = await self._client.get_all_node_info(timeout=option.timeout
)
except DataSourceUnavailable:
raise DataSourceUnavailable(GCS_QUERY_FAILURE_WARNING)
result = []
for message in reply.node_info_list:
data = self._message_to_dict(message=message,
fields_to_decode=["node_id"])
result.append(data)
result = self._filter(result, option.filters, NodeState)
# Sort to make the output deterministic.
result.sort(key=lambda entry: entry["node_id"])
return ListApiResponse(
result={d["node_id"]: d
for d in islice(result, option.limit)})
def list_jobs(self, *, option: ListApiOptions) -> ListApiResponse:
# TODO(sang): Support limit & timeout & async calls.
try:
result = []
job_info = self._client.get_job_info()
for job_id, data in job_info.items():
data = asdict(data)
data["job_id"] = job_id
result.append(data)
except DataSourceUnavailable:
raise DataSourceUnavailable(GCS_QUERY_FAILURE_WARNING)
return ListApiResponse(result=result)
async def list_runtime_envs(self, *, option: ListApiOptions) -> ListApiResponse:
"""List all runtime env information from the cluster.
Returns:
A list of runtime env information in the cluster.
The schema of returned "dict" is equivalent to the
`RuntimeEnvState` protobuf message.
We don't have id -> data mapping like other API because runtime env
doesn't have unique ids.
"""
agent_ids = self._client.get_all_registered_agent_ids()
replies = await asyncio.gather(
*[
self._client.get_runtime_envs_info(node_id, timeout=option.timeout)
for node_id in agent_ids
],
return_exceptions=True,
)
result = []
unresponsive_nodes = 0
for node_id, reply in zip(self._client.get_all_registered_agent_ids(), replies):
if isinstance(reply, DataSourceUnavailable):
unresponsive_nodes += 1
continue
elif isinstance(reply, Exception):
raise reply
states = reply.runtime_env_states
for state in states:
data = self._message_to_dict(message=state, fields_to_decode=[])
# Need to deseiralize this field.
data["runtime_env"] = RuntimeEnv.deserialize(
data["runtime_env"]
).to_dict()
data["node_id"] = node_id
result.append(data)
partial_failure_warning = None
if len(agent_ids) > 0 and unresponsive_nodes > 0:
warning_msg = NODE_QUERY_FAILURE_WARNING.format(
type="agent",
total=len(agent_ids),
network_failures=unresponsive_nodes,
log_command="dashboard_agent.log",
)
if unresponsive_nodes == len(agent_ids):
raise DataSourceUnavailable(warning_msg)
partial_failure_warning = (
f"The returned data may contain incomplete result. {warning_msg}"
)
result = self._filter(result, option.filters, RuntimeEnvState)
# Sort to make the output deterministic.
def sort_func(entry):
# If creation time is not there yet (runtime env is failed
# to be created or not created yet, they are the highest priority.
# Otherwise, "bigger" creation time is coming first.
if "creation_time_ms" not in entry:
return float("inf")
elif entry["creation_time_ms"] is None:
return float("inf")
else:
return float(entry["creation_time_ms"])
result.sort(key=sort_func, reverse=True)
return ListApiResponse(
result=list(islice(result, option.limit)),
partial_failure_warning=partial_failure_warning,
)
async def list_objects(self, *, option: ListApiOptions) -> ListApiResponse:
"""List all object information from the cluster.
Returns:
{object_id -> object_data_in_dict}
object_data_in_dict's schema is in ObjectState
"""
raylet_ids = self._client.get_all_registered_raylet_ids()
replies = await asyncio.gather(
*[
self._client.get_object_info(node_id, timeout=option.timeout)
for node_id in raylet_ids
],
return_exceptions=True,
)
unresponsive_nodes = 0
worker_stats = []
for reply, _ in zip(replies, raylet_ids):
if isinstance(reply, DataSourceUnavailable):
unresponsive_nodes += 1
continue
elif isinstance(reply, Exception):
raise reply
for core_worker_stat in reply.core_workers_stats:
# NOTE: Set preserving_proto_field_name=False here because
# `construct_memory_table` requires a dictionary that has
# modified protobuf name
# (e.g., workerId instead of worker_id) as a key.
worker_stats.append(
self._message_to_dict(
message=core_worker_stat,
fields_to_decode=["object_id"],
preserving_proto_field_name=False,
)
)
partial_failure_warning = None
if len(raylet_ids) > 0 and unresponsive_nodes > 0:
warning_msg = NODE_QUERY_FAILURE_WARNING.format(
type="raylet",
total=len(raylet_ids),
network_failures=unresponsive_nodes,
log_command="raylet.out",
)
if unresponsive_nodes == len(raylet_ids):
raise DataSourceUnavailable(warning_msg)
partial_failure_warning = (
f"The returned data may contain incomplete result. {warning_msg}"
)
result = []
memory_table = memory_utils.construct_memory_table(worker_stats)
for entry in memory_table.table:
data = entry.as_dict()
# `construct_memory_table` returns object_ref field which is indeed
# object_id. We do transformation here.
# TODO(sang): Refactor `construct_memory_table`.
data["object_id"] = data["object_ref"]
del data["object_ref"]
result.append(data)
result = self._filter(result, option.filters, ObjectState)
# Sort to make the output deterministic.
result.sort(key=lambda entry: entry["object_id"])
return ListApiResponse(
result=list(islice(result, option.limit)),
partial_failure_warning=partial_failure_warning,
)
async def list_tasks(self, *, option: ListApiOptions) -> ListApiResponse:
"""List all task information from the cluster.
Returns:
{task_id -> task_data_in_dict}
task_data_in_dict's schema is in TaskState
"""
raylet_ids = self._client.get_all_registered_raylet_ids()
replies = await asyncio.gather(
*[
self._client.get_task_info(node_id, timeout=option.timeout)
for node_id in raylet_ids
],
return_exceptions=True,
)
unresponsive_nodes = 0
running_task_id = set()
successful_replies = []
for reply in replies:
if isinstance(reply, DataSourceUnavailable):
unresponsive_nodes += 1
continue
elif isinstance(reply, Exception):
raise reply
successful_replies.append(reply)
for task_id in reply.running_task_ids:
running_task_id.add(binary_to_hex(task_id))
partial_failure_warning = None
if len(raylet_ids) > 0 and unresponsive_nodes > 0:
warning_msg = NODE_QUERY_FAILURE_WARNING.format(
type="raylet",
total=len(raylet_ids),
network_failures=unresponsive_nodes,
log_command="raylet.out",
)
if unresponsive_nodes == len(raylet_ids):
raise DataSourceUnavailable(warning_msg)
partial_failure_warning = (
f"The returned data may contain incomplete result. {warning_msg}"
)
result = []
for reply in successful_replies:
assert not isinstance(reply, Exception)
tasks = reply.owned_task_info_entries
for task in tasks:
data = self._message_to_dict(
message=task,
fields_to_decode=["task_id"],
)
if data["task_id"] in running_task_id:
data["scheduling_state"] = TaskStatus.DESCRIPTOR.values_by_number[
TaskStatus.RUNNING
].name
result.append(data)
result = self._filter(result, option.filters, TaskState)
# Sort to make the output deterministic.
result.sort(key=lambda entry: entry["task_id"])
return ListApiResponse(
result=list(islice(result, option.limit)),
partial_failure_warning=partial_failure_warning,
)
