# Reading on a stretched cluster

**URL:** <https://forum.scylladb.com/t/reading-on-a-stretched-cluster/662>\
**Category:** ScyllaDB\
**Created:** [June 29, 2023, 11:48am UTC](https://forum.scylladb.com/t/reading-on-a-stretched-cluster/662 "2023-06-29T11:48:46Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![aaponomarev](https://avatars.discourse-cdn.com/v4/letter/a/53a042/32.png) [@aaponomarev](https://forum.scylladb.com/u/aaponomarev)\
**Post date:** [June 29, 2023, 11:48am UTC](https://forum.scylladb.com/t/reading-on-a-stretched-cluster/662/1 "2023-06-29T11:48:46Z")

</div>

- We have a test stand with stretched scylladb cluster:

- the attacking host placed in dc1

- All nodes are same

- what is the reason that after the time expires, the application starts reading from host 10.0.0.3 (which is obviously located further - in another data center)

- And after some time, the situation changes exactly the opposite.

---

<div class="post-metadata">

**Author:** ![aaponomarev](https://avatars.discourse-cdn.com/v4/letter/a/53a042/32.png) [@aaponomarev](https://forum.scylladb.com/u/aaponomarev)\
**Post date:** [June 30, 2023, 8:59am UTC](https://forum.scylladb.com/t/reading-on-a-stretched-cluster/662/2 "2023-06-30T08:59:39Z")

</div>

- This night result:  
 ![screenshot-2023.06.30-11_52_32](https://us1.discourse-cdn.com/flex016/uploads/scylladb/original/1X/641f0bf9d73ee284d8bf8480fc4cbbe7bde86c3f.png)
- As can be seen from the graph: the first time reading was from the nearest nodes. But not for long 🙂 A few minutes later, a host located in another DC took the lead

---

<div class="post-metadata">

**Author:** ![Botond\_Denes](https://sea2.discourse-cdn.com/flex016/user_avatar/forum.scylladb.com/botond_denes/32/163_2.png) [@Botond\_Denes](https://forum.scylladb.com/u/Botond_Denes)\
**Post date:** [July 12, 2023, 11:17am UTC](https://forum.scylladb.com/t/reading-on-a-stretched-cluster/662/3 "2023-07-12T11:17:12Z")

</div>

When you have a multi-DC setup, make sure the Consistency Level you use for you reads is a `LOCAL_` one: `LOCAL_ONE` or `LOCAL_QUORUM`. Otherwise, the replicas will be selected from among all the nodes in the cluster, disregarding DC.  
See [Consistency Levels | ScyllaDB Docs](https://opensource.docs.scylladb.com/stable/cql/consistency.html).
