# RayOutOfMemoryError: Why is autoscaler not creating new pods?

**URL:** <https://discuss.ray.io/t/rayoutofmemoryerror-why-is-autoscaler-not-creating-new-pods/5587>\
**Category:** Kubernetes\
**Created:** [March 29, 2022, 2:20pm UTC](https://discuss.ray.io/t/rayoutofmemoryerror-why-is-autoscaler-not-creating-new-pods/5587 "2022-03-29T14:20:09Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![Dmitri](https://sea2.discourse-cdn.com/flex020/user_avatar/discuss.ray.io/dmitri/32/657_2.png) [@Dmitri](https://discuss.ray.io/u/Dmitri)\
**Post date:** [April 28, 2022, 2:59am UTC](https://discuss.ray.io/t/rayoutofmemoryerror-why-is-autoscaler-not-creating-new-pods/5587/4 "2022-04-28T02:59:24Z")

</div>

See also this thread:

> [@Tasks are completed but ray.exceptions.WorkerCrashedError](https://discuss.ray.io/t/tasks-are-completed-but-ray-exceptions-workercrashederror/5888/4):
>
> For heavy workloads, a large head node annotated with rayResources: { "CPU": 0 } is a good idea. If you’re seeing workers crashing, it might be a good idea to reduce the number of tasks that run concurrently on the workers by increasing the resource annotations on the task. For example if a task is annotated @ray.remote(num\_cpus=2) there will be at most 3 concurrent instances of that task on a CPU:6 worker node. @ray.remote is equivalent to @ray.remote(num\_cpus=1), which would result in up to…

---

_[View the full topic](https://discuss.ray.io/t/rayoutofmemoryerror-why-is-autoscaler-not-creating-new-pods/5587)._