# How much storage should be overprovisioned in cases of failure?

**URL:** <https://forum.yugabyte.com/t/how-much-storage-should-be-overprovisioned-in-cases-of-failure/771>\
**Category:** General\
**Created:** [July 19, 2020, 10:37am UTC](https://forum.yugabyte.com/t/how-much-storage-should-be-overprovisioned-in-cases-of-failure/771 "2020-07-19T10:37:01Z")\
**Posts on this page:** 1\
**Showing post:** 1

<div class="post-metadata">

**Author:** ![dorian\_yugabyte](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.yugabyte.com/dorian_yugabyte/32/206_2.png) [@dorian\_yugabyte](https://forum.yugabyte.com/u/dorian_yugabyte)\
**Post date:** [July 19, 2020, 10:37am UTC](https://forum.yugabyte.com/t/how-much-storage-should-be-overprovisioned-in-cases-of-failure/771/1 "2020-07-19T10:37:01Z")

</div>

[Question posted by a user on [YugabyteDB Community Slack](https://www.yugabyte.com/slack) ]

Assuming we have multi-DC or a multi-zone aware replica placement with RF=3, each zone has **2 (az1) + 2 (az2) + 2 (az3) = 6 nodes**. Each node has **1TB** of space.

How much storage should we over-provision on each node ?

---

_[View the full topic](https://forum.yugabyte.com/t/how-much-storage-should-be-overprovisioned-in-cases-of-failure/771)._