# Run FoundationDB cluster on multi Kuberbetes clusters

**URL:** <https://forums.foundationdb.org/t/run-foundationdb-cluster-on-multi-kuberbetes-clusters/3741>\
**Category:** Kubernetes Operator\
**Created:** [January 17, 2023, 3:19pm UTC](https://forums.foundationdb.org/t/run-foundationdb-cluster-on-multi-kuberbetes-clusters/3741 "2023-01-17T15:19:19Z")\
**Posts on this page:** 1\
**Page:** 2

<div class="post-metadata">

**Author:** ![mason](https://sea1.discourse-cdn.com/foundationdb/user_avatar/forums.foundationdb.org/mason/32/1145_2.png) [@mason](https://forums.foundationdb.org/u/mason)\
**Post date:** [February 6, 2023, 9:43am UTC](https://forums.foundationdb.org/t/run-foundationdb-cluster-on-multi-kuberbetes-clusters/3741/22 "2023-02-06T09:43:57Z")

</div>

Thank a lots for your answer.

> [@johscheuer](#):
>
> That’s not possible writes will always go to the (active) primary region and has to be committed to the primary DC and the satellite DC. A read in theory can go to every region, with the limitation, that the GRV request will always be made to the primary DC and if a version is not available in the current (remote) region the read will be blocked until the version/mutation is available or a timeout is hit. So basically the setup of multi region in FDB is an active-passive model.

Maybe this is final question 🙂  
Even I have multi\_dc cluster, read/write request always comes to primary DC, is there anyway to improve read latency? I really want read request can read data at closest DC.

[Previous page](https://forums.foundationdb.org/t/run-foundationdb-cluster-on-multi-kuberbetes-clusters/3741.md?page=1)