nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Confluence Database Location

Brandon Pal

We are currently running Confluence 4 and abotut to start evaluating Confluence 5.

When it comes to using an external database my question is does the database have to be on an secondary server or can it sit on the same server as conflunce?

Current setup:

Confluence Server

DB Server

Idea:
Conflunce & DB Server

Will this work? Is there a reason to not do this?

Thank you

Accepted answers

Matthew J. Horn

I'd recommend dumping the table data and importing it, rather than cloning the DB. Helps you separate environmental issues from data issues.

I would NOT run Jira and Confluence on the same server. Atlassian's doc recommends against it, for one, and you don't want a single point of failure.

All comments

Nic Brough -Adaptavist-

I don't think I've ever seen a corporate install that doesn't do it that way. Large organisations often have big database sets with dedicated DBA teams, with completely remote application servers and different teams to handle them.

I'd recommend making sure the database server and application server are in the same data centre (there's other stuff I'd say about it too, but it's not particularly useful without knowing your hardware, resourcing and so-on)

Brandon Pal

We are a small IT team with 3 admins.

Currently we have 1 server that has Conflunece & Jira and 1 DB server for Conflunce & JIRA.

As we wan't to test Conflunce 5 we were debating creating a conflunce & DB server in one and eventually doing the same for JIRA.

I'd like to test Conflunce 5 with our current data set but don't want to end up messing up our current version of Confluce 4 in production.

If we were to leave it as is as, I think our only option for testing 5 would be to

1) Create new VM for conflunce 5

2) clone DB VM

3) Install conflunce on new vm and connect it to DB clone.

This should allow me to test Conflunce 5 with our current data set.

From here I'm not sure what the best course of action would be?

1) Upgrade 4 to 5 on our current server

2) Create a new VM that only holds confluence and connect it to our current conflunce DB

3) Same as 2 but instead of using current db creating a new table (conflunce2) and importing the DB as a backup

Any thouhgts?

Nic Brough -Adaptavist-

You will definitely need a separate instance to test this upgrade - new application server and new database space.

I think your last option is the safest - if it goes wrong, you can simply turn off the new servers and go back to the old one.

Brandon Pal

Which would you recomend:

1) use the cloned DB as is (is that possible)

2) create a new table and import the data

In the long run would you recomend running JIRA & Conflunce on the same servers or would 2 be better?

Nic Brough -Adaptavist-

I don't know what you're talking about now. Do you mean use a clone for the test? Or on go-live? I've already said you should create a separate copy of the data and app-server., so I'm not sure what you're asking about.

I can't really recommend the same or separate servers without knowing your setup in a lot more detail. But if either comes under a heavy load, then there's a lot to be said for separate servers.

Matthew J. Horn

I'd recommend dumping the table data and importing it, rather than cloning the DB. Helps you separate environmental issues from data issues.

I would NOT run Jira and Confluence on the same server. Atlassian's doc recommends against it, for one, and you don't want a single point of failure.

Lee Anne Peders

I have just completed testing an upgrade from v3.5.5 to v5.1.5. I have more of an "enterprise" setup with separate VMs for Confluence, JIRA and the databases, but I just want to illustrate my process for testing the upgrade.

I am currently running Confluence 3.5.5 on a production VM, with the database on the production DB server. I created a new VM for the test installation with the exact disk configuration as the production VM.

I copied the <install> and <home> directories from the production server to the test server, and restored a backup of the production database to my test DB server. I very carefully edited confluence.cfg.xml to point to the test database. You DO NOT want your test instance to touch your production database. Once I confirmed that everything worked correctly, I proceeded with the upgrade testing.

As part of my production upgrade this weekend, I will be shutting down the current production VM after copying all the files to the new production VM and proceeding with the upgrade there. My backup plan will be to simply power up the old VM just in case there are issues.

Whether or not to run the tools on the same server is really going to be determined by hardware and usage. You need to test configuration and performance to make that determination for what works for your environment. I don't recommend it, if you can avoid it.

Nic Brough -Adaptavist-

Just to clarify - Atlassian's docs tell you not to run them in the same application server - the Tomcat (or other), not the physical server. Although the single point of failure is still correct of course.

Welcome!

It looks like you're new here. Sign in or register to get started.

Quick Links

Find more posts tagged with

confluence

confluence-administration