Data Center 4.0
Versions

Installing Crowd Data Center

Installation and Upgrade Guide

On this page

Related content

  • No related content found

Still need help?

The Atlassian Community is here for you.

Ask the community

To install Crowd Data Center, you'll create a cluster of Crowd instances that will make sure your users have uninterrupted access both to Crowd, and all other systems that are connected to it. We recommend that you start with the prerequisite information, listed on this page, to understand what Data Center is, and to know exactly what you'll need to complete the installation.

 

Before you begin

Before you install Crowd Data Center, you need to answer a few questions.

 

What is Crowd Data Center?

Tell me more...

You should first understand what Crowd Data Center is, and how it works. We've gathered some resources that will help you get to know the high-level overview of Data Center, and its architecture.

Take a look at Crowd Data Center.

How do I get it?
Tell me more...

You'll need two things to get started – an installer, and a license. There's no special installer for Data Center – you just install Crowd, and then enable Data Center features in it.

As for the license, get your Crowd Data Center license, or try it out for free.

What are the prerequisites?

Tell me more...

Supported platforms

Supported operating systems, databases, etc., are the same as for the Server installation, and you can see them here: Supported platforms. You need to use an external database - HSQLDB is not supported.

Node requirements

Those specific to Data Center include requirements for nodes that create the cluster:

  • Each node is a separate machine (physical or virtual). They don't need to be identical, but should be as similar as possible for consistent performance.
  • All nodes are running the same version of Crowd. You'll be copying Crowd from one node to another, so this shouldn't be a problem.
  • They use the same timezone, and have the current time synced. You can use ntpd to set this up.
  • All nodes share a common database, also installed on a separate machine.
  • All nodes can access the shared home directory. You can set it up using NFS, or a similar solution. We'll mention it in this guide.
Do I need a load balancer?
Tell me more...

Yes. Crowd Data Center relies on a load balancer to balance the traffic between the nodes, and this guide assumes that you already have one set up. You can use a load balancer of your choice, just make sure it meets these requirements:

  • Supports "cookie based session affinity", also known as "sticky sessions".
  • Can route HTTP/HTTPS traffic to one of the available nodes.
  • Can determine whether a node is available or not, and route requests to other nodes if needed.
  • All Atlassian applications and other REST clients must access your nodes through the load balancer.

Or you can just turn your proxy into a load balancer.

Many bigger installations of Crowd already have a reverse proxy configured, and many reverse proxies can do load balancing as well. We've provided some examples on how to use your proxy as a load balancer. See Load balancer examples.

1. Install Crowd

Crowd Data Center is available for Crowd 3.0, or later. If you're not on this version yet, install or upgrade your Crowd instance. 

Crowd installation and upgrade guide

After you've installed Crowd and applied your license, you can verify that Data Center is available by going to  > Licensing.

2. Set up the shared directory

You'll need to create a remote directory that is readable and writable by all nodes in the cluster. There are multiple ways to do this, but the simplest is to use an NFS share.

  1. Stop Crowd.

  2. Go to Crowd's home directory, and check whether it already has the shared sub-directory, which might have been created after starting Crowd. If it's there, you'll need to copy its contents to the new shared directory that you'll create in the next steps. If you can't find it, just omit this step.
  3. Create a remote directory, accessible by all nodes in the cluster, and name it shared. 

  4. Mount shared as a sub-directory of the Crowd home directory.

    <home-directory>/shared

3. Add the first Crowd node to your load balancer

The load balancer distributes the traffic between the nodes. If a node stops working, the remaining nodes will take over its workload, and your users won't even notice it.

  1. First, add your load balancer as a trusted proxy server in Crowd. See Configuring Trusted Proxy Servers.

  2. Add the first node to the load balancer.

  3. Restart the node, and then try opening different pages in Crowd. If the load balancer is working properly, you should have no problems with accessing Crowd.

  4. In Crowd, go to > Clustering. The node should be listed as part of the cluster. 

    tip/resting Created with Sketch.

    If your load balancer supports health checks, configure it to perform a check on http://<crowd-node>:8095/<context-path>/status, where <crowd-node> is the node's hostname or IP address, and <context-path> is the Crowd's context path (e.g. /crowd). If the node doesn't respond with a 200 OK response within a reasonable time, the load balancer shouldn't direct any traffic to this node.

  5. After you've added the node to the load balancer, configure the Crowd's base URL to also point to the load balancer. Go to > General, and enter the URL of your load balancer as Base URL.

4. Add the remaining nodes to the cluster

  1. Copy the Crowd installation directory to the new node. 
  2. Create a home directory, like you did for the first node, and mount shared as its sub-directory.

  3. Edit crowd-init.properties, and enter the path to the home directory that you just created. 

    Where can I find crowd-init.properties?

    The crowd-init.properties file is in <installation-directory>\crowd-webapp\WEB-INF\classes\.

  4. Go to <installation-directory>/apache-tomcat/conf/Catalina/localhost, and delete the openidserver.xml file. This is needed because currently the CrowdID component doesn't support clustering, and it must be enabled only on the first node. The component will work as usual.
  5. Start Crowd. It will read the configuration from the shared directory, and start without any extra setup.
  6. Take a look around the new Crowd instance. Verify that user and group management, directory synchronization, and any custom integrations all work as expected.
  7. Again, verify that the node was added to the cluster. In Crowd, go to  > Clustering.

     

  8. If everything looks fine, you can configure your load balancer to start routing traffic to the new node. Once you do this, you can make a couple of changes in one Crowd instance to see if they're visible in other instances as well.

What else?

Adding node names

When displaying information about your nodes in the Crowd footer or on the  > Clustering page, Crowd Data Center uses random IDs that were generated for your nodes. Instead, you can give them more persistent and readable names by setting the cluster.node.name system property, like in the following example:

CATALINA_OPTS=-Dcluster.node.name=node-1


Well done! Crowd Data Center is now at your service.


Interested in learning more about what Crowd Data Center provides? Click here for an overview.

Last modified on Mar 30, 2020

Was this helpful?

Yes
No
Provide feedback about this article

Related content

  • No related content found
Powered by Confluence and Scroll Viewport.