Move node to different cluster
Hi Guys, I have 4 node in DC cluster and 2 node in DRC cluster, i want to move one of the node from DC to DRC so i have equal resources on both sites. Can you share the best practices to minimalize downtime? Also i don't seem to see the option to remove hci nodes from the cluster. Hi,moving a node from one Sangfor HCI cluster to another involves several steps and careful planning to minimize downtime. Here are some general steps you can follow:Data Backup: Initiate the process by ensuring that a recent backup of all vital data on the node to be moved is available.
VM Migration: Transfer all Virtual Machines (VMs) from the node to be moved to other nodes within the same cluster. This can be accomplished by exporting the VM and subsequently importing it to another HCI.
Cluster Controller Change: If the node to be moved is currently a cluster controller, you’ll need to assign another node as the cluster controller. This can be done by selecting the desired node and then choosing More > Set to Cluster Controller.
Node Removal: Once the VMs have been migrated and the cluster controller has been reassigned (if required), the node can be removed from the cluster. Please note that the Sangfor HCI interface does not directly provide the option to remove HCI nodes from the cluster. It is advisable to seek assistance from Sangfor support for this step.
Node Addition to New Cluster: After successful removal of the node from the old cluster, it can be added to the new cluster. Follow the standard procedure for adding a node to a Sangfor HCI cluster.
Resource Rebalancing: Once the node has been added to the new cluster, rebalance the resources as required to ensure optimal performance.
Sangfor Support Consultation: It’s important to remember to consult with Sangfor support throughout this process. They have the necessary expertise and tools to ensure a safe and controlled migration process while maintaining data integrity and service continuity. Follow these steps;
(i) Ensure data replication is up-to-date between clusters.
(ii) Safely power down the node to be moved in DC.
(iii) Physically relocate the node to DRC and power it up.
(iv) Verify connectivity and rejoin the node to the DRC cluster. first step is both nodes are updated with latest logs.
then down the main site
then manully up the DRS
then check the connectivity
To minimize downtime when moving a node from the DC to the DRC cluster in your HCI setup, follow these best practices: Firstly, ensure data replication is up-to-date between the clusters. Next, safely power down the node being moved and document its configuration. Remove the node from the DC cluster management interface and ensure any data services running on the node are redirected to other nodes. Configure the moved node in the DRC cluster with the appropriate network settings and add it to the cluster. Validate the node's status and perform failover tests before resuming normal operations. Hi,
If your nodes are in production, i advise to contact Sangfor support to help you with this work and to have an backup plan if anything goes wrong.
thanks, Last edited by Newbie290036 07 May 2024 23:17.
Moving a node from a primary data center (DC) cluster to a disaster recovery center (DRC) cluster in a hyper-converged infrastructure (HCI) environment requires careful planning to minimize downtime and ensure a seamless transition. Please follow below steps to proceed:
[*]Assess Resource Availability: Before proceeding, evaluate the resource utilization and workload distribution across both clusters to ensure that moving a node from the DC to the DRC will indeed balance resources effectively.
[*]Data Redundancy and Backup: Ensure that there is sufficient data redundancy and backup in place before initiating the migration. This ensures data integrity and availability during the migration process and minimizes the risk of data loss.
[*]Plan Maintenance Window: Schedule the migration during a planned maintenance window to minimize disruption to ongoing operations. Coordinate with stakeholders to select a time when workload impact is minimal, such as during off-peak hours.
[*]Live Migration: Whenever possible, utilize live migration capabilities offered by your HCI platform. Live migration enables you to move the workload from the node scheduled for relocation to other nodes in the DC cluster without interrupting service. Once the workload is migrated, you can proceed with relocating the physical node to the DRC.
[*]Network Considerations: Ensure that the network connectivity between the DC and DRC clusters is robust and sufficient to handle the migration process. Consider factors such as bandwidth, latency, and network security to ensure smooth data transfer between clusters.
[*]Failover and Failback Procedures: Prepare failover and failback procedures in advance to handle any unexpected issues during the migration. Test these procedures in a controlled environment to verify their effectiveness and identify any potential issues before the actual migration.
[*]Documentation and Communication: Document the migration plan, including step-by-step instructions, configuration settings, and rollback procedures. Communicate the migration plan to all relevant stakeholders, including IT teams, end-users, and management, to ensure everyone is aware of the upcoming changes and potential impact.
[*]Post-Migration Verification: After completing the migration, perform thorough testing to verify the functionality and performance of the migrated workload in the DRC cluster. Monitor system performance, network connectivity, and application functionality to ensure everything is functioning as expected.
Hi,
Directly moving an HCI node from one cluster (DC) to another (DRC) with minimal downtime needs strategic planning. HCI clusters are designed for HA, and removing a node usually requires a full decommission and reconfiguration on the new cluster.
You may try to consider the approaches below and see if this will suffice your requirement.
1. VMs Migration - most common approach to achieve resource balance. You can try migrating VMs from the overloaded DC cluster to the underutilized DRC cluster. Most HCI solutions offer built-in live migration capabilities that minimize downtime while transferring VMs which keeps the original cluster structure intact.
2. Standby VM approach - Consider setting up a standby VM in the DRC cluster pre-configured with critical applications. If a failure occurs in the DC cluster, you can quickly failover to the standby VM in the DRC with minimal downtime. This needs additional configuration but assures faster recovery for critical services.
3. Building new node in DRC - If resource imbalance is a long-term issue, consider adding a new node directly to the DRC cluster which will provide a more permanent solution and avoids the complexities of moving existing nodes.
4. Decommissioning and Rebuilding (highest downtime and least recommended) - you could decommission the node in the DC cluster, rebuild it with the DRC cluster configuration, and then migrate all the VMs to the newly added node. This has the most downtime but allows a complete shift of resources. HiHi,
Please find the attached files for Sangfor HCI Disaster Recovery Solutions for your reference.
I hope they can help you resolve your inquiries. Thank you.
Moving an HCI node from your datacenter (DC) cluster to your disaster recovery (DRC) cluster can be done with minimal downtime by following these best practices:
Preparation:
Validate the DRC cluster: Ensure your DRC cluster is healthy and has enough storage and compute resources to accommodate the additional node.
Live Migration (if applicable): Check if your HCI solution supports live migration of virtual machines (VMs) between clusters. This allows VMs to be transferred to other nodes while they remain operational.
Planned downtime window: Schedule a maintenance window for the move to minimize disruption to running workloads.
Drain the node: Stop any VMs or services running on the node you plan to move. This minimizes data movement during the process.
Moving the Node:
The specific steps for removing a node will depend on your HCI software. Here's a general guideline:
Windows Admin Center (for Azure Stack HCI):
Open Windows Admin Center and select Cluster Manager.
Connect to your DC cluster.
Go to Tools > Servers.
Select the node you want to remove and click Remove.
PowerShell (for various HCI solutions):
Consult your HCI software's documentation for the specific PowerShell command to remove a node. It might involve commands like Disable-ClusterNode or Remove-ClusterNode.
Once the node is removed from the DC cluster, add it to the DRC cluster using the appropriate method for your software.
Post-Migration:
Verify node health: Ensure the newly added node is functioning correctly in the DRC cluster.
Start VMs: If you live-migrated VMs, verify they are running smoothly on their new hosts.
Test failover (optional): Consider performing a controlled failover to the DRC cluster to validate your disaster recovery setup.
page:
[1]
2