Fix bug from the previous fixes.
Add more tests
Stop using m5.xlarge (not supported now)
There are 2 hard blockers from the infra: 1. Large size disk is not supported. 2. m5.xlarge is not supported. Both are considered as a high priority to be fixed soon.
The `py_modules` field of runtime_env supports uploading local Python modules for use on the Ray cluster. One gap in this is if the local Python module is in the form of a wheel (`.whl` file.) This PR adds the missing support for uploading and installing the `.whl` file.
The Ray Dashboard starts Serve in the `"_ray_internal_dashboard"` namespace. However, Serve by default starts in the `"serve"` namespace. This causes surprising behavior when working with the Serve CLI and REST API.
This change make the Ray Dashboard start Serve in the `"serve"` namespace, allowing the REST API to work intuitively with the Python API.
The current multi node tests use a hardcoded mapping for local development mounts. With this PR, a new environment variable is introduced to be able to control this dynamically. Additionally, some minor improvements to the test utilities and monitor are added.
GCS pubsub has been the default for awhile. There is little chance that we would need to revert back to Redis pubsub in future. This is the step in removing Redis pubsub, by first removing the `enable_gcs_pubsub()` feature guard.
In this PR, we disabled every 10min FullGC which is not triggered by a global gc event in Java worker. As detail, we added `triggered_by_global_gc` flag to indicate whether the gc event is triggered by a global gc event. If it's triggered by global gc, we still need to do FullGC.
Co-authored-by: Qing Wang <jovany.wq@antgroup.com>
Adds a streaming based reading option for Snappy-compressed files. Arrow doesn't support streaming Snappy decompression since the canonical C++ Snappy library doesn't natively support streaming decompression. This PR works around this by doing streaming reads of snappy-compressed files using the streaming decompression API provided in the [python-snappy](https://github.com/andrix/python-snappy) package.
This commit supplies a custom datasource that uses Arrow + [python-snappy](https://github.com/andrix/python-snappy) to read and decompress Snappy-compressed files.
Co-authored-by: siddharth.goel <siddharth.goel@bytedance.com>
Co-authored-by: Chen Shen <scv119@gmail.com>
This PR adds a feature that allows user to make their training runs more reproducible. I've implemented this feature by following PyTorch's guide on how to limit sources of randomness (https://pytorch.org/docs/stable/notes/randomness.html).
These changes will make it easier for us to benchmark Ray Train, and also make it easier for users to reproduce their experiments.
`Application` stores a group of deployments and can write them to a YAML config. However, this requires the deployments to use import paths as their `func_or_class`. This change make all deployments in an `Application` store only import paths as the `func_or_class`.
This change also adds a utility function to get a deployment's import path. This utility function is used in the DeploymentNode for Pipelines.
Interface for DataParallelTrainer and updates to ScalingConfig definition.
Depends on #22986
Co-authored-by: Eric Liang <ekhliang@gmail.com>
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
Co-authored-by: matthewdeng <matthew.j.deng@gmail.com>
Currently, when a spill/restore worker fails and the state of it in the worker pool is idle, the worker pool will not clean up the metadata of the worker. Subsequent spill/restore requests will reuse this dead worker and RPC requests cannot succeed. This results in broken object spilling functionality.
This PR addresses the issue by removing disconnected IO workers from `registered_io_workers` and `idle_io_workers`.
Previously DatasetPipeline stages were executed by one actor each, which compromised fault tolerance through lineage reconstruction. This centralizes all task submission at the pipeline coordinator to improve fault tolerance. To preserve pipeline parallelism, the stages are executed by a threadpool. To clean up the threadpool, the pipeline coordinator adds any running threads to a global set that is checked by the threads during `ray.wait`.
Note that this will only provide fault tolerance for split pipes if all pipeline consumers stay alive. It will not work if one of the consumers dies and restarts because next_dataset_if_ready is not idempotent.
Co-authored-by: Clark Zinzow <clarkzinzow@gmail.com>
Adds the following interfaces (without implementation, for discussion / approval):
- `serve.Application`
- `serve.DeploymentNode`
- `serve.DeploymentMethodNode`, `serve.DAGHandle`, and `serve.drivers.PipelineDriver`
- `serve.run` & `serve.build`
In addition to these Python APIs, we will also support the following CLI commands:
- `serve run [--blocking=true] my_file:my_node_or_app # Uses Ray client, blocking by default.`
- `serve build my_file:my_node output_path.yaml`
- `serve deploy [--blocking=false] # Uses REST API, non-blocking by default.`
- `serve status [--watch=false] # Uses REST API, non-blocking by default.`
Adds an ability for users to specify a custom results post-processing function that will be applied to metrics before they are reported to Tune in XGBoost/LightGBM integration callbacks, allowing for support for xgb.cv/lgbm.cv. Updates example to show it in action and in CI.
