TorchX Schedulers define plugins to existing schedulers. Used with the runner, they submit components as jobs onto the respective scheduler backends. TorchX supports a few schedulers out-of-the-box. You can add your own by implementing .. py:class::torchx.schedulers and registering it in the entrypoint.
- torchx.schedulers.get_schedulers(session_name: str, **scheduler_params: object) Dict[str, Scheduler] ¶
get_schedulers returns all available schedulers.
- torchx.schedulers.get_scheduler_factories() Dict[str, SchedulerFactory] ¶
get_scheduler_factories returns all the available schedulers names and the method to instantiate them.
The first scheduler in the dictionary is used as the default scheduler.
- class torchx.schedulers.Scheduler(backend: str, session_name: str)¶
An interface abstracting functionalities of a scheduler. Implementors need only implement those methods annotated with
- cancel(app_id: str) None ¶
Cancels/kills the application. This method is idempotent within the same thread and is safe to call on the same application multiple times. However when called from multiple threads/processes on the same app the exact semantics of this method depends on the idempotency guarantees of the underlying scheduler API.
This method does not block for the application to reach a cancelled state. To ensure that the application reaches a terminal state use the
- close() None ¶
Only for schedulers that have local state! Closes the scheduler freeing any allocated resources. Once closed, the scheduler object is deemed to no longer be valid and any method called on the object results in undefined behavior.
This method should not raise exceptions and is allowed to be called multiple times on the same object.
Override only for scheduler implementations that have local state (
torchx/schedulers/local_scheduler.py). Schedulers simply wrapping a remote scheduler’s client need not implement this method.
- abstract describe(app_id: str) Optional[DescribeAppResponse] ¶
Describes the specified application.
AppDef description or
Noneif the app does not exist.
- log_iter(app_id: str, role_name: str, k: int = 0, regex: Optional[str] = None, since: Optional[datetime] = None, until: Optional[datetime] = None, should_tail: bool = False, streams: Optional[Stream] = None) Iterable[str] ¶
Returns an iterator to the log lines of the
k``th replica of the ``role. The iterator ends end all qualifying log lines have been read.
If the scheduler supports time-based cursors fetching log lines for custom time ranges, then the
untilfields are honored, otherwise they are ignored. Not specifying
untilis equivalent to getting all available log lines. If the
untilis empty, then the iterator behaves like
tail -f, following the log output until the job reaches a terminal state.
The exact definition of what constitutes a log is scheduler specific. Some schedulers may consider stderr or stdout as the log, others may read the logs from a log file.
Behaviors and assumptions:
Produces an undefined-behavior if called on an app that does not exist The caller should check that the app exists using
exists(app_id)prior to calling this method.
Is not stateful, calling this method twice with same parameters returns a new iterator. Prior iteration progress is lost.
Does not always support log-tailing. Not all schedulers support live log iteration (e.g. tailing logs while the app is running). Refer to the specific scheduler’s documentation for the iterator’s behavior.
- 3.1 If the scheduler supports log-tailing, it should be controlled
Does not guarantee log retention. It is possible that by the time this method is called, the underlying scheduler may have purged the log records for this application. If so this method raises an arbitrary exception.
should_tailis True, the method only raises a
StopIterationexception when the accessible log lines have been fully exhausted and the app has reached a final state. For instance, if the app gets stuck and does not produce any log lines, then the iterator blocks until the app eventually gets killed (either via timeout or manually) at which point it raises a
should_tailis False, the method raises
StopIterationwhen there are no more logs.
Need not be supported by all schedulers.
Some schedulers may support line cursors by supporting
iterseeks to the 50th log line).
- Whitespace is preserved, each new line should include
support interactive progress bars the returned lines don’t need to include
\nbut should then be printed without a newline to correctly handle
- Whitespace is preserved, each new line should include
streams – The IO output streams to select. One of: combined, stdout, stderr. If the selected stream isn’t supported by the scheduler it will throw an ValueError.
Iteratorover log lines of the specified role replica
NotImplementedError – if the scheduler does not support log iteration
- run_opts() runopts ¶
Returns the run configuration options expected by the scheduler. Basically a
- abstract schedule(dryrun_info: AppDryRunInfo) str ¶
submitexcept that it takes an
AppDryRunInfo. Implementors are encouraged to implement this method rather than directly implementing
submitcan be trivially implemented by:
dryrun_info = self.submit_dryrun(app, cfg) return schedule(dryrun_info)
- submit(app: AppDef, cfg: T, workspace: Optional[str] = None) str ¶
Submits the application to be run by the scheduler.
WARNING: Mostly used for tests. Users should prefer to use the TorchX runner instead.
The application id that uniquely identifies the submitted app.
- submit_dryrun(app: AppDef, cfg: T) AppDryRunInfo ¶
Rather than submitting the request to run the app, returns the request object that would have been submitted to the underlying service. The type of the request object is scheduler dependent. This method can be used to dry-run an application. Please refer to the scheduler implementation’s documentation regarding the actual return type.
- class torchx.schedulers.SchedulerFactory(*args, **kwargs)¶