Confluence Server Emergency Concept

Question

Hello guys,

I would like to implement a Confluence Emergency Concept in my company , right now we are running Confluence Server with the application on Windows(VM) and DB on AIX(Oracle).

My question and some of my requirement are :

- Available physical on location (other location) if the data center is gone

- Offline available not dependent of AD

Andrei Trif · Answer

Do you have any other suggestions ?

Andrei Trif · Answer

First off all thank you Nic for the detailed answer.

But to give you a more detailed overview of what I already have in mind.  I will like to share my idea .

The server I want will be a physical machine and in the case of a disaster where we don’t have the data centre anymore due to floods, earthquake, fire or maybe natural disasters I still have a running Confluence on this PC, also I don’t need it to be ready and accessible for all the users in the company, this sever will be accessed only by some IT admins to see the some crucial documentation.

In this hypothetical case , they will see the last synchronized documentation.

This is why I ask the community , can I use this scenario like :

- PC with Windows server 2016

- Data base sync/copy from AIX machine (Oracle ) to this Windows PC (the frequency and full/incremented method will be determined later)

Question here : Can I have the DB on the same computer in this scenario?

- Copy/Sync of the “Files store” folder

- No need to have AD sync for the user, I will use a local Admin account to access the articles.

Nic Brough -Adaptavist- · Answer

The standard minimal setup for an emergency standby like this is

* A copy of the server on a remote site, with Confluence installed, but not running.
* Database replication set up so that production data is replicated over to the emergency site.
* Attachments mounted on shared storage, so there is a copy of production on both sides.

When it goes bang, you turn off (what is left of) Production, fire up the emergency site, trigger re-indexing and redirect the network to look at it instead of production.

There's lots of refining you could (and should) do to that, and it's not quite as simple as I've said as you'll need to fiddle with some settings to make it all work the way you expect, and think about how to go back to production later.  But it's the basics.