Creation of Proxmox HA cluster

This article describes the configuration of a High Availability Cluster or HA Cluster using three servers with Proxmox VE 8.2.2. In this example, Ceph is configured within the cluster as the data storage.

Requirements for High Availability

Before configuring the HA cluster, it must be verified if all requirements are fulfilled:

• 3 server systems identical in design
• all nodes must be located in the same system
• date and time must be synchronized (NTP)
• SSH must be released for SSH on port 22

Installation and configuration

First, Proxmox must be installed on all 3 nodes. Please make sure that you have entered the correct host name and IP configuration.

Network configuration

After that, the network configuration of all nodes must be made. The following steps must be executed after the installation has been successful:

• 
  First, select the host name and click on [Network]. Next, you will see the initial network configuration, which can be adjusted according to your requirements.
• 
  After that, the changes are applied by clicking on [Apply]. This process must be repeated on the remaining nodes.

Now, the network configuration can be verified with a ping test. You can also test the ping with jumbo frames, if configured. The following command would test this with an MTU configured to 9000 bytes:

ping -M do -s 8972 [IP-Adresse]

Create Proxmox VE cluster

The following steps must be executed to create the cluster:

• 
  Go to [Datacenter] [Cluster] and click on [Create Cluster]. Now, a pop-up appears where a Clustername and the Network are entered for the PVE-cluster. By clicking on [Create] , the cluster is created.
• 
  Now you can copy the join information by selecting the [Join Information] option and clicking [Copy Information]. This step is required for adding the remaining nodes.
• 
  To add the remaining nodes to the cluster, connect to their GUI and under [Datacenter] [Cluster], click [Join Cluster]. After that, the join information, the password and the network can be entered. By clicking on [Join 'Clustername'], the host is added. Repeat this process for the remaining servers.
• 
  The overview should look similar to this.

Hint for a cluster with a lot of nodes: For clusters with a large number of nodes (25 or more), enter all nodes in the local hosts file on each cluster node. Otherwise, it may come to high network traffic and stability problems.

Add Ceph storage

To create a Ceph storage for storing all data, the following steps must be performed:

• 
  Select a node here and click on [Ceph] and [Install Ceph] to start the installation.
• 
  After the installation has been completed, the Ceph-Network can be stated. In addition, you can also set the number of replicates here. Click on [Next] to complete the configuration. This process is repeated for all nodes.
• 
  Next, some services must be configured for Ceph. The first are the monitoring and management services. For this, click [Ceph], then [Monitor] and then [Create] in Monitor and Manager. In this example, one service is created per host.
• 
  The result should look similar to this image.
• 
  The HDDs/SSDs/NVMes must be prepared for the creation of the OSDs. To do this, delete them under [Disks] [Wipe Disk] and initialize them with GPT [Initialize Disk with GPT]. Repeat this with all data carriers.
• 
  Now, you can create the OSDs under [Ceph] [OSD] by clicking on [Create: OSD]. Select the disk and a device class and confirm it by clicking on [Create]. This must be repeated for all data carriers.
• 
  After creating the OSD, a Meta Data server can be created under [CephFS]. These are required for CephFS. For this, click on [Create]. In this example, a MDS host is created per host.
• 
  You can now create CephFS by clicking [Create CephFS]. You can use the default values here.
• 
  Finally, a pool must be created to store all data from the VMs/CTs. For this, switch to [Pools] and click [Create]. Next, you must assign a pool name; you can also set a target ratio of 0.9.
• 
  Finally, you can check the Ceph dashboard, and everything should be showing as Healthy.

Configuration High Availability

Now you can create the required virtual machines and containers on the cluster. To take advantage of the cluster's high availability, it is essential that the configured VMs/CTs are added to the HA. To set this up, follow the steps below:

• 
  Go to [Datacenter] and click on [HA]. You can now add the VM to HA by clicking [Add]. To do so, enter the VM ID and a request state, then confirm by clicking [Add].

After all steps have been executed, the cluster is functional and your virtual instance is highly available.

Author: Niklas Pauli Niklas Pauli works in the product management team at Thomas-Krenn. He completed his general university entrance qualification at Freyung High School and then his training as an IT specialist for system integration at Thomas-Krenn.AG. After completing his training, he continued to work in the areas of Proxmox VE incl. Ceph, backup (Proxmox, Veeam, SEP). He is also a regular writer for the Thomas-Krenn-Wiki.

Author: Tim Lindinger Tim Lindinger has been working for Thomas-Krenn in Technical Support since 2015. His areas of responsibility include the setup of test systems, the repair of RMA servers as well as the support for customer inquiries. On the software side Tim has specialized in the topic Proxmox VE.

Translator: Alina Ranzinger Alina has been working at Thomas-Krenn.AG since 2024. After her training as multilingual business assistant, she got her job as assistant of the Product Management and is responsible for the translation of texts and for the organisation of the department.