Is restoring JIRA from an XML backup or dbdump.sql is the same thing?

Question

Hi,

I have installed a local instance of JIRA, and I have an XML backup and dbdump.sql from my production server. 
I want to make the exact replica of production server.

So, my question is, if restoring from an XML backup and Dbdump.sql is the same thing? 
If not, then which would make the exact replica of production server (including the database and everything)

Regards,
Tayyab

Nic Brough -Adaptavist- · Answer

Totally different things, although you will end up with the same system if you follow the steps correctly.

* The easy way to imagine this is that
* You have a database that acts as JIRA's data store
* You can copy that out as a dump, suitable for your flavour of database
* You can use JIRA's backup routine that reads the data store and dumps out the whole thing as xml
* If you restore the database, you get an exact copy
* If you restore the xml, using exactly the same version of JIRA, you get an exact copy of the original (barring bugs in the importers in JIRA), although you may find some things are accidentally cleaned out by the process

However, on their own, NEITHER method is a full backup.  They are just the raw JIRA data.

You also need to think about the attachments, AND, if you are thinking about things like disaster recovery, you're going to need to think about the system configuration files too - the stuff that makes up your actual JIRA installation.

Josh Steckler · Answer

Both could essentially produce the same result, but it would be a different set of steps.

The xml option helps you walk through it more, for people with less DBA experience (because it assumes you have a working database already). It will also rebuild your index if you didn't copy that across, prompts you for a license, and allows you to disable mail via a radio button. But you can do all of this yourself with the db dump.

In the end, in my experience it's a question of which you're more comfortable with. There are some valid reasons why the db dump may be better, described here: https://confluence.atlassian.com/adminjiraserver073/backing-up-data-861253815.html