* [runtime env] runtime env inheritance refactor (#22244)
Runtime Environments is already GA in Ray 1.6.0. The latest doc is [here](https://docs.ray.io/en/master/ray-core/handling-dependencies.html#runtime-environments). And now, we already supported a [inheritance](https://docs.ray.io/en/master/ray-core/handling-dependencies.html#inheritance) behavior as follows (copied from the doc):
- The runtime_env["env_vars"] field will be merged with the runtime_env["env_vars"] field of the parent. This allows for environment variables set in the parent’s runtime environment to be automatically propagated to the child, even if new environment variables are set in the child’s runtime environment.
- Every other field in the runtime_env will be overridden by the child, not merged. For example, if runtime_env["py_modules"] is specified, it will replace the runtime_env["py_modules"] field of the parent.
We think this runtime env merging logic is so complex and confusing to users because users can't know the final runtime env before the jobs are run.
Current PR tries to do a refactor and change the behavior of Runtime Environments inheritance. Here is the new behavior:
- **If there is no runtime env option when we create actor, inherit the parent runtime env.**
- **Otherwise, use the optional runtime env directly and don't do the merging.**
Add a new API named `ray.runtime_env.get_current_runtime_env()` to get the parent runtime env and modify this dict by yourself. Like:
```Actor.options(runtime_env=ray.runtime_env.get_current_runtime_env().update({"X": "Y"}))```
This new API also can be used in ray client.
This PR adds precise reason details regarding worker failures. All information is available either by
- ray list workers
- exceptions from actor failures.
Here's an example when the actor is killed by a SIGKILL (e.g., OOM killer)
```
RayActorError: The actor died unexpectedly before finishing this task.
class_name: G
actor_id: e818d2f0521a334daf03540701000000
pid: 61251
namespace: 674a49b2-5b9b-4fcc-b6e1-5a1d4b9400d2
ip: 127.0.0.1
The actor is dead because its worker process has died. Worker exit type: UNEXPECTED_SYSTEM_EXIT Worker exit detail: Worker unexpectedly exits with a connection error code 2. End of file. There are some potential root causes. (1) The process is killed by SIGKILL by OOM killer due to high memory usage. (2) ray stop --force is called. (3) The worker is crashed unexpectedly due to SIGSEGV or other unexpected errors.
```
## Design
Worker failures are reported by Raylet from 2 paths.
(1) When the core worker calls `Disconnect`.
(2) When the worker is unexpectedly killed, the socket is closed, raylet reports the worker failures.
The PR ensures all worker failures are reported through Disconnect while it includes more detailed information to its metadata.
## Exit types
Previously, the worker exit types are arbitrary and not correctly categorized. This PR reduces the number of worker exit types while it includes details of each exit type so that users can easily figure out the root cause of worker crashes.
### Status quo
- SYSTEM ERROR EXIT
- Failure from the connection (core worker dead)
- Unexpected exception or exit with exit_code !=0 on core worker
- Direct call failure
- INTENDED EXIT
- Shutdown driver
- Exit_actor
- exit(0)
- Actor kill request
- Task cancel request
- UNUSED_RESOURCE_REMOVED
- Upon GCS restart, it kills bundles that are not registered to GCS to synchronize the state
- PG_REMOVED
- When pg is removed, all workers fate share
- CREATION_TASK (INIT ERROR)
- When actor init has an error
- IDLE
- When worker is idle and num workers > soft limit (by default num cpus)
- NODE DIED
- Only can detect when the node of the owner is dead (need improvement)
### New proposal
Remove unnecessary states and add “details” field. We can categorize failures by 4 types
- UNEXPECTED_SYSTEM_ERROR_EXIT
- When the worker is crashed unexpectedly
- Failure from the connection (core worker dead)
- Unexpected exception or exit with exit_code !=0 on core worker
- Node died
- Direct call failure
- INTENDED_USER_EXIT.
- When the worker is requested to be killed by users. No workflow required. Just correctly store the state.
- Shutdown driver
- Exit_actor
- exit(0)
- Actor kill request
- Task cancel request
- INTENDED_SYSTEM_EXIT
- When the worker is requested to be killed by system (without explicit user request)
- Unused resource removed
- Pg removed
- Idle
- ACTOR_INIT_FAILURE (CREATION_TASK_FAILED)
- When the actor init is failed, we fate share the process with the actor.
- Actor init failed
## Limitation (Follow up)
Worker failures are not reported under following circumstances
- Worker is failed before it registers its information to GCS (it is usually from critical system bug, and extremely uncommon).
- Node is failed. In this case, we should track Node ID -> Worker ID mapping at GCS and when the node is failed, we should record worker metadata.
I will create issues to track these problems.
There is a race condition where the failure message is sent but resubscription hasn't been established. In this case, we'll lose this message.
This PR fixes the test case only by making sure resubscription is done before killing the worker.
We are not going to fix this issue for the first version of GCS HA because:
1. Right now, only failed worker will be reported and it'll be used to eagerly kill the unnecessary workers. From correctness, it's ok.
2. worker/task support is not P0 for GCS HA (Ray Serve). Actor doesn't have this issue since the subscription for the actor is by actor id and raylet will fetch the actor status when resubscribing.
Things are not working:
- we won't kill workers which are not useful anymore. So for a very extreme case (the worker hangs), it'll have a resource leak.
This fixes two bugs in data locality:
When a dependent task is already in the CoreWorker's queue, we ran the data locality policy to choose a raylet before we added the first location for the dependency, so it would appear as if the dependency was not available anywhere.
The locality policy did not take into account spilled locations.
Added C++ unit tests and Python tests for the above.
Related issue number
Fixes#24267.
This PR adds a FAQ to Datasets docs.
Docs preview: https://ray--24932.org.readthedocs.build/en/24932/
## Checks
- [x] I've run `scripts/format.sh` to lint the changes in this PR.
- [x] I've included any doc changes needed for https://docs.ray.io/en/master/.
- [x] I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/
- Testing Strategy
- [x] Unit tests
- [ ] Release tests
- [ ] This PR is not tested :(
Co-authored-by: Eric Liang <ekhliang@gmail.com>
Previously we pinned the controller on the head node because it was the SPOF, now with GCS HA work that will no longer be the case so it should be able to be run on any node. We still prefer the head node for non-HA cases.
This PR adds a dedicated docs page for examples, and adds a basic e2e tabular data processing example on the NYC taxi dataset.
The goal of this example is to demonstrate basic data reading, inspection, transformations, and shuffling, along with ingestion into dummy model trainers and doing dummy batch inference, for tabular (Parquet) data.
This PR overhauls the "Accessing Datasets", adding proper coverage of each data consuming methods, including the ML framework exchange APIs (to_torch() and to_tf()).
It seems `_InactiveRpcError` is not saved as the last exception. Raise an explicit error when publishing fails after retries.
For log and error publishing, dropping messages should be tolerable so log the exception instead.
The previous implementation of TorchPredictor failed when the model outputted an array for each sample (for example outputting logits). This is because Pandas Dataframes cannot hold numpy arrays of more than 1 element. We have to convert the outermost numpy array returned by the model to a list so that it be can stored in a Pandas Dataframe.
The newly added test fails under the old implementation.
For mac and windows, we default to start the head node with 127.0.0.1 to avoid security pop-ups and assume people won't use cluster mode a lot. However, when people do want to create a mac/windows cluster, our connection message is confusing since we cannot connect to the head node that's listening on 127.0.0.1. This PR tells people what to do in this case (e.g. use node-ip-address).
This PR adds a fast path for `sync_dir_between_nodes` that gets triggered if both source IP and target IP are the same. It uses simple `shutil` operations instead of packing and unpacking to improve performance.
This PR moves all exception classes from runtime module to api module. It's aiming to eliminate the confusion about ray exceptions. It means that Ray users don't need to touch runtime module when API programming after this PR.
Note that this should be merged onto 2.0.
The documentation says
> Consider using .map_batches() for better performance (you can implement filter by dropping records).
but there aren't any examples of how to do so.
This PR knocks off a few miscellaneous GA docs P0s given in our docs tracker. Namely:
- Documents Datasets resource allocation model.
- De-emphasizes global/windowed shuffling.
- Documents lazy execution mode, and expands our execution model docs in general.
I have seen errors where the prometheus view data dictionary is changed when iterating over it. So make a copy (which is atomic) before iterating.
Also, use a separate thread for GCS client heartbeat RPC. This avoids the issue of missing heartbeat when raylet is too busy.
This PR supports various output formatting. By default, we support yaml format. But this can be changed depending on the UX research we will conduct in the future.
1cb0f4b51a5799e0360a66db01000000:
actor_id: 1cb0f4b51a5799e0360a66db01000000
class_name: A
state: ALIVE
f90ba34fa27f79a808b4b5aa01000000:
actor_id: f90ba34fa27f79a808b4b5aa01000000
class_name: A
state: ALIVE
Table format is not supported yet. We will support this once we enhance the API output (which I will create an initial API review soon).
This PR adds a flag to enable creating nodes in the main thread in the autoscaler. The flag is turned on for the KubeRay node provider. KubeRay already uses a flag to disable node updater threads -- with the changes from this PR, the autoscaler becomes single-threaded when launching on KubeRay.
This is part of the Dataset GA doc fix effort to update/improve the documentation.
This PR revamps the Getting Started page.
What are the changes:
- Focus on basic/core features that are bread-and-butter for users, leave the advanced features out
- Focus on high level introduction, leave the detailed spec out (e.g. what are possible batch_types for map_batches() API)
- Use more realistic (yet still simple) data example that's familiar to people (IRIS dataset in this case)
- Use the same data example throughout to make it context-switch free
- Use runnable code rather than faked
- Reference to the code from doc, instead of inlining them in the doc
Co-authored-by: Ubuntu <ubuntu@ip-172-31-32-136.us-west-2.compute.internal>
Co-authored-by: Eric Liang <ekhliang@gmail.com>
When an object is under reconstruction, pull manager keeps the bundle request active with no timeout, which may block the next bundle request that's needed for the object reconstruction. As a result, we have deadlock.
For example, task 1 takes object A as argument and returns object B, task 2 takes object B as argument. When we run task 2, pull manager will add B to the queue and then B is lost. In this case, task 1 is re-submitted and A is added the the pull manager queue after B (assuming both tasks are scheduled to the same node). Due to limited available object store memory, A cannot be activated until B is pulled but B cannot be pulled until A is pulled and B is reconstructed.
The solution is that if an active pull request has pending-creation objects, pull manager will deactivates it until creation is done. This way, we will free object store memory occupied by the current active pull request so that next requests can proceed and potentially unblock the object creation.
This is the 1st PR to remove the code path of multiple core workers in one process. This PR is aiming to remove the flags and APIs related to `num_workers`.
After this PR checking in, we needn't to consider the multiple core workers any longer.
The further following PRs are related to the deeper logic refactor, like eliminating the gap between core worker and core worker process, removing the logic related to multiple workers from workerpool, gcs and etc.
**BREAK CHANGE**
This PR removes these APIs:
- Ray.wrapRunnable();
- Ray.wrapCallable();
- Ray.setAsyncContext();
- Ray.getAsyncContext();
And the following APIs are not allowed to invoke in a user-created thread in local mode:
- Ray.getRuntimeContext().getCurrentActorId();
- Ray.getRuntimeContext().getCurrentTaskId()
Note that this PR shouldn't be merged to 1.x.
This example simply doesn't run as is. We can bring it back up again later, if it makes sense. But it's not clear what the variables used there, like actor are. Fixes#21328
Signed-off-by: Max Pumperla <max.pumperla@googlemail.com>
as per discussion on slack, this should avoid having old versions of the docs indexed by search engines, see readthedocs/readthedocs.org#2430 for reference.
Signed-off-by: Max Pumperla <max.pumperla@googlemail.com>
This PR is a general overhaul of the "Creating Datasets" feature guide, providing complete coverage of all (public) dataset creation APIs and highlighting features and quirks of the individual APIs, data modalities, storage backends, etc. In order to keep the page from getting too long and keeping it easy to navigate, tabbed views are used heavily.
Instead of letting Datasets implicitly use cluster resources in the margins of explicit allocations of other libraries, such as Tune, Datasets should provide an option for explicitly allocating resources for a Datasets workload for users that want to box Datasets in. This PR adds such an explicit resource allocation option, via exposing a top-level scheduling strategy on the DatasetContext with which a placement group can be given.
