Relocation for Sangfor HCI

LuisKuan96 Lv1Posted May-22-2025 15:12

Last edited by LuisKuan96 2025-May-22 15:15.

Hey everyone, I have questions for relocation of SANGFOR HCI. We have a set of HCI Cluster consist of few nodes, running in our existing Data Center1. This set HCI Cluster connected to SCP in another set of HCI Cluster in Data Center2 . Recently, we going to move/relocate Data Center1 SANGFOR HCI to new datacenter, Data Center3. The IP segment at new datacenter is totally different with the existing one(Data Center1). What is the best practice steps/procedures to proceed it?

AimanHakim has solved this question and earned 10 coins.

Posting a reply earns you 2 coins. An accepted reply earns you 20 coins and another 10 coins for replying within 10 minutes. (Expired) What is Coin?

Enter your mobile phone number and company name for better service. Go

Last edited by AimanHakim 2025-May-23 04:09.

(In reply to LuisKuan96)
In that case, here's what I propose to do. I'm around 90% sure it will work even though I've never done it before but I've done like using standalone licensing as its base. What I've done was in the same data center (Let's call it data center 1), the client wanted DR features so they have 2 clusters, we'll call it Cluster-A n Cluster-B.
Cluster-A is on HQ while the Cluster-B is meant for DR. Due to worry of slow transmission via WAN causing the process to be slow n time consuming, Cluster-B was deployed on HQ temporarily so it can somewhat directly using same internal network for DR process done faster. After completion, Cluster-B is disconnected SCP and from HQ network and I configured it to the DR network segment. Then the Cluster-B relocated to DR site and connected it back to the SCP. It works normally.

The part I'm not sure yet as I don't have the official proof is the licensing validity after disconnect from SCP. When cluster is centralised and the cluster suddenly don't have access or can't ping to SCP, it will automatically have a grace period 30 days to renew the license. This is what I found out during my test lab. I used SCP trial license to test the centralised licensing. SCP install on Cluster-C and assign the license to Cluster-D. Once the SCP license expires, I thought the Cluster-D also expired but somehow it got that 30 days grace period to be operational. So the actual enterprise license should work the same way as the trial license. I'll try double check with the FAE team about this later.

So ur situation is somewhat similar to mine as both are required to be bind to SCP in order to work. So here's my guide for ya...

Step 1
- You can just shutdown the all the VMs normally. Followed by delete the Cluster IP. Then change the nodes IP based on the network segment in data center 3. Do bear in mind *DO NOT CHANGE THE NODE IP WHICH IS THE CLUSTER CONTROLLER FIRST*. What I mean is that make sure u change the other nodes first. The node IP which is the cluster controller will be configured last once the other node's IP are done changed. If you accidently changed the node IP which is the cluster controller first, it will unable to ping the other nodes. It will cause the other nodes POV see the cluster controller node as offline and vice versa. When it does happen, u won't be able to configure the other nodes. Even if you direct connect to other nodes or configured via KVM over IPMI, u won't be able to change the IP of the other nodes as it detects the cluster controller is offline. Below is the example, in this case there are 2 nodes in my cluster. Change the controller's node IP 1st as shown in the blue circle. Then lastly change the cluster controller's IP.



Once done, you can do a connection test via nodes itself either back 2 back or via switch. You'll be able to see the cluster is online and the nodes are connected and online with network segment of data center 3. Then you may proceed turn off the nodes and ready for the relocation.

Step 2
- Once relocated just plug in normally and powered on the servers. Once the nodes are up, configure the new cluster IP based on the network segment in data center 3.

Step 3
- Then log in to the SCP located in data center 2. Go to clusters option and reconfigured the cluster's old IP with the new one. Ofcourse make sure the cluster IP is able to be reachable to SCP. If ur data center 3 is already reachable to the SCP in data center 2, it will be no issue.

Step 4
- Once the cluster status is back online, the centralised license will sync back and the cluster in data center 3 is operational. Ofcourse the VM inside of it require to change it's IP based on the network segment data center 3.
Is this answer helpful?
Doll Lv1Posted Jun-01-2025 01:18
  
Relocating your Sangfor HCI cluster from Data Center 1 to Data Center 3 requires careful planning due to the different IP segments. Here's a step-by-step guide to help you through the process:

*Pre-Migration Steps*

- *Assess and Document*: Document your current HCI cluster setup, including network configurations, IP addresses, and connections to the SCP in Data Center 2.
- *Plan IP Addressing*: Plan the new IP addressing scheme for Data Center 3, ensuring it doesn't conflict with existing IP addresses in Data Center 2.

*Migration Steps*

- *Update Network Configurations*: Update the network configurations on your Sangfor HCI nodes to use the new IP addresses for Data Center 3. This may involve reconfiguring virtual IP addresses, subnet masks, gateways, and DNS settings.
- *Configure Routing*: Ensure proper routing between Data Center 3 and Data Center 2, allowing communication between the HCI clusters and the SCP.
- *Test Connectivity*: Verify connectivity between the HCI clusters and the SCP after updating the network configurations.

*Post-Migration Steps*

- *Verify Cluster Functionality*: Validate that the HCI cluster is functioning correctly in Data Center 3, including storage, computing, and networking capabilities.
- *Test Disaster Recovery*: If you have disaster recovery configurations, test them to ensure they work correctly between Data Center 2 and Data Center 3.
- *Update Documentation*: Update your documentation to reflect the new IP addresses and network configurations for the HCI cluster in Data Center 3.

*Additional Considerations*

- *Sangfor Support*: Reach out to Sangfor's technical support for guidance on the migration process, especially if you're unsure about specific configurations or encounter issues.
- *Change Management*: Schedule the migration during a maintenance window to minimize downtime and ensure a smooth transition.
- *Testing*: Perform thorough testing after the migration to ensure all systems are functioning as expected ¹.
Newbie102711 Lv1Posted May-22-2025 19:56
  
Best Steps to Move SANGFOR HCI to a New Data Center with Different IPs
If you're moving your SANGFOR HCI cluster from Data Center1 to a new Data Center3 where the IP addresses are completely different, here’s a simple way to handle it:

1. Prepare and Backup
First, make a list of all your nodes, current IP addresses, VLANs, and storage setups. Don’t forget to back up all important VMs and system settings before you start. Also, plan out the new IP addresses you’ll use at the new data center for management, storage, and business networks.

2. Shut Down and Export
Gracefully shut down all your VMs. If you want to be extra safe, export important VMs separately. Then, power off the HCI nodes so you can move them safely.

3. Physical Move
Take the hardware to the new data center, rack it, and connect all power cables, network cables, and storage connections properly.

4. Change IP Settings
Once everything is set up in the new location, log into each node directly (using a monitor or management port) and change their IP addresses to match the new network setup. Make sure all nodes can communicate with each other after changing the IPs.

5. Bring the Cluster Back Online
If the cluster doesn’t automatically start working together, you might need to manually rejoin the nodes to the cluster using their new IP addresses. Also, update any SCP settings in the other data center to recognize the new IPs.

6. Test Everything
Turn on some test VMs and check if everything is working—network, storage, performance—make sure all services are running smoothly.

7. Final Checks
Update your documentation with the new IPs and configurations, and keep an eye on the system to catch any issues early. It’s also a good idea to set up logging and alerts to prevent problems in the future.
AimanHakim Lv2Posted May-22-2025 15:34
  
Hi, does the cluster in data center 1 is using centralized licensed provided by the SCP in data centre 2 or standalone but using SCP for DR purposes etc?

I Can Help:

Change

Board Leaders

NyxZale...

Weekly Sharers

Newbie5...

Weekly Questioners