# What's the policy of auto-scaling in Ray streaming?

**URL:** <https://discuss.ray.io/t/whats-the-policy-of-auto-scaling-in-ray-streaming/6560>\
**Category:** Ray Data\
**Created:** [June 20, 2022, 11:03am UTC](https://discuss.ray.io/t/whats-the-policy-of-auto-scaling-in-ray-streaming/6560 "2022-06-20T11:03:56Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Ting\_Sun](https://sea2.discourse-cdn.com/flex020/user_avatar/discuss.ray.io/ting_sun/32/2765_2.png) [@Ting\_Sun](https://discuss.ray.io/u/Ting_Sun)\
**Post date:** [June 20, 2022, 11:03am UTC](https://discuss.ray.io/t/whats-the-policy-of-auto-scaling-in-ray-streaming/6560/1 "2022-06-20T11:03:57Z")

</div>

**How severe does this issue affect your experience of using Ray?**

- None: Just asking a question out of curiosity

---

<div class="post-metadata">

**Author:** ![Ting\_Sun](https://sea2.discourse-cdn.com/flex020/user_avatar/discuss.ray.io/ting_sun/32/2765_2.png) [@Ting\_Sun](https://discuss.ray.io/u/Ting_Sun)\
**Post date:** [June 20, 2022, 11:04am UTC](https://discuss.ray.io/t/whats-the-policy-of-auto-scaling-in-ray-streaming/6560/2 "2022-06-20T11:04:59Z")

</div>

Will it support DS2 from OSDI 2018? [https://www.usenix.org/system/files/osdi18-kalavri.pdf](https://www.usenix.org/system/files/osdi18-kalavri.pdf)

---

<div class="post-metadata">

**Author:** ![Ting\_Sun](https://sea2.discourse-cdn.com/flex020/user_avatar/discuss.ray.io/ting_sun/32/2765_2.png) [@Ting\_Sun](https://discuss.ray.io/u/Ting_Sun)\
**Post date:** [June 20, 2022, 11:06am UTC](https://discuss.ray.io/t/whats-the-policy-of-auto-scaling-in-ray-streaming/6560/3 "2022-06-20T11:06:09Z")

</div>

Or using external metrics such as CPU/memory utilizations?

---

<div class="post-metadata">

**Author:** ![jianxiao](https://avatars.discourse-cdn.com/v4/letter/j/e480ec/32.png) [@jianxiao](https://discuss.ray.io/u/jianxiao)\
**Post date:** [August 1, 2022, 11:15pm UTC](https://discuss.ray.io/t/whats-the-policy-of-auto-scaling-in-ray-streaming/6560/4 "2022-08-01T23:15:46Z")

</div>

Current streaming solution is Mobius ([GitHub - ray-project/mobius: Mobius is an AI infra platform including realtime computing and training.](https://github.com/ray-project/mobius)), which I don’t think have the support of your linked paper.