Microsoft Windows allows you to configure a failover cluster. As part of this configuration a single cluster IP address can be configured. This IP address is hosted on the master node of the cluster and if the master node fails, it is automatically failed over to the new master node of the cluster.
As can be seen in the documentation, on OCI, a vNIC can have one or more IP addresses assigned to it. It will always have a primary IP address and can have multiple secondary IP addresses. The IP address of each node in the cluster is assigned as the primary IP address of the vNIC assigned to that node. The cluster IP address is assigned as a secondary IP address, initially to the vNIC of the master node in the cluster. When the master node fails, whilst the Windows clusterware will fail the cluster IP address to the new master node automatically within the cluster, the secondary IP address also has to be reassigned to the vNIC of this new master node. The script provided in the repo performs that functionality.
The script supports a 2 node failover cluster for Windows Server 2016 or Windows Server 2019 with an optional DR node in a different data center. It was designed to provide failover for nodes running SQL Server and optionally you can configure a SQL Server cluster name and IP in which case the script will failover the IP for this cluster as well as for the default Windows failover cluster.
- A knowledge of Windows Server setup and administration is required.
- A knowledge of Oracle Cloud Infrastructure and use of the OCI Python SDK is required.
- A 2 node Windows Failover Cluster must be setup and configured on OCI before using the script, and the node and cluster IP addresses defined in OCI. Optionally a SQL Server cluster can also be setup.
- A directory in an identical location of your choice on both nodes.
- The script is written in Python and uses the OCI Python SDK.
- Both must be installed on both nodes of the windows cluster prior to installing and configuring the script.
On both nodes of the windows cluster:
- Copy the contents of the repo to the directory previously created
- Edit oci-mscluster-scheduler.xml and replace with the path to the Python 3 binary on each Windows host and the path to the directory where you have copied the repo in the working directory
- Edit the oci_config file. It is preformatted. It has placed holders for the values required to configure the OCI Python SDK, these are documented here. Please note add the real path for "key_file". The script will use the oci_config file in preference to any config file you have already configured for the OCI Python SDK.Edit oci_config and provide the Tenancy ID, fingerprint and key_file of something unspecified in your documentation. I need to go somewhere else to find what this is.
- Edit settings.json as per the table below
| Field Name | Description |
|---|---|
| node1_name | The name of the first node that is configured in the Windows Cluster |
| node2_name | The name of the second node that is configured in the Windows Cluster |
| sql_cluster_name | If SQL Server is in a cluster as well, specify the configured cluster name. If not specified private_ip_id_sql_cluster will be ignored and is not required. |
| private_ip_id_default_cluster | OCID of the secondary IP address configured as the Windows cluster IP address |
| private_ip_id_sql_cluster | OCID of the secondary IP addeess configured as the SQL Server IP address |
| skip_dr_node_name | If the cluster contains a DR node in a different DataCenter, specify its name here and it will be ignored by the script (which is for HA not DR) |
| vnic_1 | OCID of the primary vNIC of Node 1 |
| vnic_2 | OCID of the primary vNIC of Node 2 |
- Add the script to Task Scheduler by using the oci-mscluster-scheduler.xml file:
eg:schtasks.exe /create /tn IPFailover /xml "<repo directory>oci-mscluster-scheduler.xml" /f
- The Issues page of this GitHub repository.