# Replication

**URL:** <https://discourse.chef.io/t/replication/1908>\
**Category:** Chef Infra (archive)\
**Created:** [June 3, 2011, 3:05pm UTC](https://discourse.chef.io/t/replication/1908 "2011-06-03T15:05:12Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Tristan\_Sloughter](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.chef.io/tristan_sloughter/32/819_2.png) [@Tristan\_Sloughter](https://discourse.chef.io/u/Tristan_Sloughter)\
**Post date:** [June 3, 2011, 3:05pm UTC](https://discourse.chef.io/t/replication/1908/1 "2011-06-03T15:05:12Z")

</div>

I have a Chef server that I need to move to a new physical server. From what  
I understand replication/clustering is suppose to be somewhat simple in Chef  
but I can’t find documentation on this.

I could get away in this instance with simply trying to manually copy over  
/var/lib/couchdb and /etc/chef to the new server but it would be nice to use  
this as a test for future cases where replication would be more appropriate.

Is there any documentation on this?

Thanks,  
Tristan

---

<div class="post-metadata">

**Author:** ![Brian\_Akins](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.chef.io/brian_akins/32/298_2.png) [@Brian\_Akins](https://discourse.chef.io/u/Brian_Akins)\
**Post date:** [June 3, 2011, 7:22pm UTC](https://discourse.chef.io/t/replication/1908/2 "2011-06-03T19:22:46Z")

</div>

On Fri, Jun 3, 2011 at 11:05 AM, Tristan Sloughter \<  
[tristan.sloughter@gmail.com](mailto:tristan.sloughter@gmail.com)\> wrote:

> I could get away in this instance with simply trying to manually copy over  
> /var/lib/couchdb and /etc/chef to the new server but it would be nice to use  
> this as a test for future cases where replication would be more appropriate.

We do "hot/warm" redundancy.

We replicate couch using normal couch replication. We use rsync for the  
cookbooks (/var/chef/checksums). We rebuild the solr index once an our or  
so on our "warm" box. The chef servers are managed by chef, so /etc/chef is  
handled that way.