ContextQMD
Back to Scylladb

Create a ScyllaDB Cluster - Multi Data Centers (DC)

docs/operating-scylla/procedures/cluster-management/create-cluster-multidc.rst

latest8.3 KB
Original Source

Create a ScyllaDB Cluster - Multi Data Centers (DC)

.. _create-cluster-multi-config-table:

Consult with the table below if each node in the cluster has internal IP for internal DC communication and external IP for cross DC communication.

Single Multi Data Centers Configuration Table

===================== =========================== Parameter Multi DC ===================== =========================== seeds External IP address

listen_address Internal IP address

rpc_address Internal IP address

broadcast_address External IP address

broadcast_rpc_address External IP address

endpoint_snitch GossipingPropertyFileSnitch ===================== ===========================

Important

If the node has two physical network interfaces in a multi-datacenter installation:

  • Set listen_address to this node's private IP or hostname.
  • Set broadcast_address to the second IP or hostname (for communication between data centers).
  • Set listen_on_broadcast_address to true.
  • Open the storage_port or ssl_storage_port on the public IP firewall.

Prerequisites

.. include:: /operating-scylla/procedures/_common/create_cluster_prerequisites_include.rst

  • Decide the name of the rack, for example, RACK1, RACK2 or RC1, RC2.
  • Decide the name of the data center, for example, DC1, DC2 or US-DC, ASIA-DC.

Choose the data center name carefully. It is not possible to rename a data center later

When working with production environments, you must choose one of the snitches below:

  • :ref:Ec2multiregionsnitch <snitch-ec2-multi-region-snitch> - for AWS cloud-based, multi-data-center deployments.

  • :ref:GossipingPropertyFileSnitch <snitch-gossiping-property-file-snitch> - for bare metal and cloud (other than AWS) deployments.

Procedure

  1. Install ScyllaDB on the nodes you want to add to the cluster. See :doc:Getting Started</getting-started/index> for further instructions, create as many nodes that you need. Follow the ScyllaDB install procedure up to scylla.yaml configuration phase.

In case that your node starts during the process follow :doc:these instructions </operating-scylla/procedures/cluster-management/clear-data>

  1. On each node, edit the scylla.yaml file to configure the parameters listed below. The file can be found under /etc/scylla/.
  • cluster_name - Set the selected cluster_name
  • seeds - Specify the IP of the node you chose to be a seed node. New nodes will use the IP of this seed node to connect to the cluster and learn the cluster topology and state.
  • listen_address - IP address that the ScyllaDB use to connect to other ScyllaDB nodes in the cluster
  • endpoint_snitch - Set the selected snitch
  • rpc_address - Address for CQL client connection
  1. In the cassandra-rackdc.properties file, edit the rack and data center information. The file can be found under /etc/scylla/.

To save bandwidth, add the prefer_local=true parameter. ScyllaDB will use the node private (local) IP address when the nodes are in the same data center.

  1. Start the nodes.

.. include:: /rst_include/scylla-commands-start-index.rst

  1. Verify that the nodes have been added to the cluster using
    nodetool status.

For Example:

In this example, we will show how to install a nine nodes cluster.

  1. Install nine ScyllaDB nodes, three nodes in each data center (U.S, ASIA, EUROPE). The IP's are:

.. code-block:: shell

U.S Data-center Node# Private IP Public IP Node1 192.168.1.201 54.187.36.59 (seed) Node2 192.168.1.202 54.187.142.201 Node3 192.168.1.203 54.187.168.20

ASIA Data-center Node# Private IP Public IP Node4 192.168.1.204 54.191.72.56 Node5 192.168.1.205 54.187.25.99 Node6 192.168.1.206 54.191.2.121

EUROPE Data-center Node# Private IP Public IP
Node7 192.168.1.207 54.160.174.243 Node8 192.168.1.208 54.235.9.159 Node9 192.168.1.209 54.146.228.25

  1. In each ScyllaDB node, edit the scylla.yaml file. See :ref:Single Multi Data Centers Configuration Table <create-cluster-multi-config-table> for reference.

U.S Data-center - 192.168.1.201

.. code-block:: shell

cluster_name: 'multi_dc_demo' seeds: "54.187.36.59" endpoint_snitch: GossipingPropertyFileSnitch rpc_address: "192.168.1.201" listen_address: "192.168.1.201" broadcast_address: "54.187.36.59" broadcast_rpc_address: "54.187.36.59" listen_on_broadcast_address: true (optional)

ASIA Data-center - 192.168.1.204

.. code-block:: shell

cluster_name: 'multi_dc_demo' seeds: "54.187.36.59" endpoint_snitch: GossipingPropertyFileSnitch rpc_address: "192.168.1.204" listen_address: "192.168.1.204" broadcast_address: "54.191.72.56" broadcast_rpc_address: "54.191.72.56" listen_on_broadcast_address: true (optional)

EUROPE Data-center - 192.168.1.207

.. code-block:: shell

cluster_name: 'multi_dc_demo' seeds: "54.187.36.59" endpoint_snitch: GossipingPropertyFileSnitch rpc_address: "192.168.1.207" listen_address: "192.168.1.207" broadcast_address: "54.160.174.243" broadcast_rpc_address: "54.160.174.243" listen_on_broadcast_address: true (optional)

  1. In each ScyllaDB node, edit the cassandra-rackdc.properties file with the relevant rack and data center information

Nodes 1-3

.. code-block:: shell

dc=US-DC rack=RACK1

Nodes 4-6

.. code-block:: shell

dc=ASIA-DC rack=RACK1

Nodes 7-9

.. code-block:: shell

dc=EUROPE-DC rack=RACK1

  1. Start the nodes.

.. include:: /rst_include/scylla-commands-start-index.rst

  1. Verify that the nodes have been added to the cluster by using the nodetool status command.

.. code-block:: shell

nodetool status

Datacenter: US-DC

Status=Up/Down |/ State=Normal/Leaving/Joining/Moving -- Address Load Tokens Owns Host ID Rack UN 54.191.2.121 120.97 KB 256 ? c84b80ea-cb60-422b-bc72-fa86ede4ac2e RACK1 UN 54.191.72.56 109.54 KB 256 ? 129087eb-9aea-4af6-92c6-99fdadb39c33 RACK1 UN 54.187.25.99 104.94 KB 256 ? 0540c7d7-2622-4f1f-a3f0-acb39282e0fc RACK1

Datacenter: ASIA-DC

Status=Up/Down |/ State=Normal/Leaving/Joining/Moving -- Address Load Tokens Owns Host ID Rack UN 54.160.174.243 109.54 KB 256 ? c7686ffd-7a5b-4124-858e-df2e61130aaa RACK1 UN 54.235.9.159 109.75 KB 256 ? 39798227-9f6f-4868-8193-08570856c09a RACK1 UN 54.146.228.25 128.33 KB 256 ? 7a4957a1-9590-4434-9746-9c8a6f796a0c RACK1

Datacenter: EUROPE-DC

Status=Up/Down |/ State=Normal/Leaving/Joining/Moving -- Address Load Tokens Owns Host ID Rack UN 54.187.36.59 114.35 KB 256 ? 4c3e1533-1b78-45bf-8bd4-818090f019ab RACK1 UN 54.187.142.201 109.54 KB 256 ? d99967d6-987c-4a54-829d-86d1b921470f RACK1 UN 54.187.168.20 109.54 KB 256 ? 2329c2e0-64e1-41dc-8202-74403a40f851 RACK1

Preventing Quorum Loss

If your cluster is symmetrical, i.e., it has an even number of datacenters with the same number of nodes, consider adding an arbiter DC to mitigate the risk of losing a quorum at a minimum cost. See :doc:Preventing Quorum Loss in Symmetrical Multi-DC Clusters </operating-scylla/procedures/cluster-management/arbiter-dc> for details.

See also:

:doc:Create a ScyllaDB Cluster - Single Data Center (DC) </operating-scylla/procedures/cluster-management/create-cluster>