Comments (3)
Hi @abatilo, thank you for opening the issue. You may have some misunderstanding for GCS FT. You can read https://docs.ray.io/en/latest/cluster/kubernetes/user-guides/kuberay-gcs-ft.html#kuberay-gcs-ft for more details. Currently, the only use case of GCS FT is for Ray Serve high availability.
from kuberay.
Thanks for responding @kevin85421. My original question still stands then. Is there no way to have the jobs running in progress survive after the Ray head gets restarted? I see that there's a way to resume a trainer from previous state but do I always have to re-submit that if the Ray head is gone?
from kuberay.
My current understanding is that Ray Train provides some degree of fault tolerance.
- If a Ray worker Pod crashes, Ray Train will launch new Ray tasks or actors, allowing the job to continue running.
- If the Ray head crashes, the driver process, which is typically running on the Ray head of the Ray job, will also crash. Therefore, the Ray tasks or actors will also be garbage collected automatically. In most cases, Ray tasks or actors are fate-sharing with the driver process. Only detached actors do not share this fate with the driver.
- In this case, users need to write the fault tolerance logic at the application level, specifically in their Ray Python script.
if (checkpoint exists): read checkpoint else: start from scratch train the model
- In this case, users need to write the fault tolerance logic at the application level, specifically in their Ray Python script.
The long-running Ray job is currently on our roadmap of the next release. We are currently working on:
- Retry mechanism in both RayJob CRD /
ray job submit
. - Best practice for checkpointing.
If you are interested in this topic, you can reach out to me on Ray Slack (my handle is 'Kai-Hsun Chen (ray team)'). We can discuss your requirements and ensure there are no feature gaps for your use cases.
from kuberay.
Related Issues (20)
- [Bug] no feedback about failure to create submitter pod due to invalid spec HOT 2
- Good repository HOT 1
- [Bug] What's minimum permission set for kuberay-operator? HOT 3
- [Feature] Jupyter ecosystem support on kuberay HOT 4
- [Bug] Can't scaler up when using autoscaler v2 HOT 6
- [Bug] ray cluster getHeadServiceIp failed if "app.kubernetes.io/name" set HOT 2
- [Feature] keep the RayCluster's prefix in the worker pods name HOT 6
- [Feature] REP 54: Implement the ReplicaFailures condition HOT 1
- [CI] Check the consistency between role.yaml and multiple_namespaces_role.yaml HOT 1
- [Feature] REP 54: Donβt assign the rayv1.Failed to the State field HOT 2
- [Refactor] REP 54: Refactor usages of the inconsistentRayClusterStatus function HOT 2
- [Feature] REP 54: Add PodName to the HeadInfo HOT 1
- [Feature] REP 54: Implement the HeadReady condition HOT 3
- [Feature] REP 54: Re-define rayv1.Ready HOT 3
- [Bug] spec.workerGroupSpecs[0].template.spec.containers[0].env: Invalid value: "null" HOT 3
- [Feature] Add API reference documentation for KubeRay custom resources HOT 5
- [Feature] Event record for failed Pod creation HOT 4
- [Bug] Leader Election Lost: Kuberay pod restarts every 5mins! HOT 26
- [Feature] Add a strict parser for YAMLs in CI HOT 1
- Adding --proxy=off for wget health check
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
π Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. πππ
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google β€οΈ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from kuberay.