20. Demystifying Virtual SAN (vSAN) Architecture & Components| SDS, Hybrid vs. All-Flash | Use Cases
Vložit
- čas přidán 8. 08. 2024
- 20. vSphere 7.x_ Virtual SAN (vSAN) Architecture & Home Lab
vSphere Storage - Overview
Software-Defined Storage (SDS) Concepts
Virtual SAN (vSAN) Architecture
vSAN Components
vSAN Hybrid vs. All-Flash
vSAN Disk Groups
vSAN Hardware Requirements
vSAN Datastore Objects
vSAN VM Storage Policies
vSAN Use cases
vSAN Datastore Creation & Management - Home lab
Prepare for vSAN Cluster Setup
Create a Cluster and Enable vSAN
Using Quick-Start Method
Manual Method
Validate the vSAN Datastore Configuration
vSAN Cluster Health Status
Verify VM on the vSAN Datastore
#VMware
#vsphere
#ServerFundamentals
#Virtualization
#VCP-DCV
#ESXi
#VM
#Networking
#vCenter
#VCSA
#VCSA Deployment
#vCenter
#Architecture
#SDDC
#VAMI
#NSX-T
#VSAN
#VRO
#VRA
#Block Storage
#File Storage
#Object Storage
#Storage
vSphere Storage
#FC SAN
#iSCSI SAN
#FCoE SAN
#vSAN
#vVols
#Virtual Volumes
#Virtual SAN
#Multipathing
#NFS 3
#NFS 4.1
#real time scenarios
#ThankYou
Please refer to the following playlist for your review.
Gnan Cloud Garage Playlists
www.youtube.com/@gnancloudgar...
VMware vSphere 7 & VMware vSphere Plus (+) | Data Center Virtualization
• VMware vSphere |VCP - ...
vSphere 7.x - Home lab - Quick Bytes | Data Center Virtualization
• vSphere 7.x - Home lab...
VMware vSphere 8
• VMware vSphere 8
VMware vSAN 8
• VMware vSAN 8
VMware NSX 4.x | Network Virtualization
• VMware NSX 4.0.0.1 | N...
VMware Cloud Foundation (VCF)+
• VMware Cloud Foundatio...
VMware Aria Automation (formerly, vRealize Automation) | Unified Multi-Cloud Management
• VMware Aria Automation...
Interview Preparation for Technical Consultants, Systems Engineers & Solution Architects
• Interview Preparation ...
VMware Tanzu Portfolio | Application Modernization
• VMware Tanzu Portfolio...
Modern Data Protection Solutions
• Modern Data Protection...
Storage, Software-Defined Storage (SDS)
• Storage, Software-Defi...
Zerto, a Hewlett Packard Enterprise (HPE) Company
• Zerto, a Hewlett Packa...
The Era of Multi-Cloud Services|HPE GreenLake Solutions|Solution Architectures|Solution Designs
• The Era of Multi-Cloud...
Gnan Cloud Garage (GCG) - FAQs |Tools |Tech Talks
• Gnan Cloud Garage (GCG...
VMware Aria Operations (formerly, vROps)
• VMware Aria Operations...
PowerShell || VMware PowerCLI
• PowerShell || VMware P...
Hewlett Packard Enterprise (HPE) Edge to Cloud Solutions & Services
• Hewlett Packard Enterp...
DevOps || DevSecOps
• DevOps || DevSecOps
Red Hat Openshift Container Platform (RH OCP)
• Red Hat Openshift Cont...
Windows Server 2022 - Concepts
• Windows Server 2022, 2...
Red Hat Enterprise Linux (RHEL) 9 - Concepts
• Red Hat Enterprise Lin...
Microsoft Azure Stack HCI
• Microsoft Azure Stack HCI
NVIDIA AI Enterprise
• NVIDIA AI Enterprise
Gratitude | Thank you messages
• Gratitude | Thank you ...
Explained in very easy simple terminology and all understand the concept..Thanks for all your effort.
You are most welcome
20. vSphere 7.x_ Virtual SAN (vSAN) Architecture & Home Lab
vSphere Storage - Overview
Software-Defined Storage (SDS) Concepts
Virtual SAN (vSAN) Architecture
vSAN Components
vSAN Hybrid vs. All-Flash
vSAN Disk Groups
vSAN Hardware Requirements
vSAN Datastore Objects
vSAN VM Storage Policies
vSAN Use cases
vSAN Datastore Creation & Management - Home lab
Prepare for vSAN Cluster Setup
Create a Cluster and Enable vSAN
Using Quick-Start Method
Manual Method
Validate the vSAN Datastore Configuration
vSAN Cluster Health Status
Verify VM on the vSAN Datastore
Lot of knowledge you share thankyou sir...
It's my pleasure Amarnath :-)
Thank YOu for detailed explanation very useful.
Glad it was helpful!
excellent teaching
Thank you
Please post for nsx the way explaination is peak bro keep posting
Sure, I will do that Bro. Thank you :-)
really thank you very much,
you are awesome
thanks a lot for sharing knowledge
Hi,
Thank you so much for your kind words!
I'm glad that you find value in the content I'm sharing on the "Gnan Cloud Garage" CZcams channel 🙂
Thanks again for watching and supporting my channel!
Fantastic
Thank you Santhosh:-)
Excellent presentation and delivery. May i use your example myself to presentation. please let me know.
Yes, Please Go ahead! Thank you :-)
Hi Sir. What are the Hardware requirements for configuring vSAN cluster? How much CPU, Memory, and Storage are required? How much storage is required for the Cache tier and capacity Tier? Kindly answer this question, sir. thanks
Hi Sir,
To set up a vSAN cluster, there are specific hardware specifications we need to meet.
Here’s a comprehensive breakdown of the requirements:
Hardware Specifications for vSAN Cluster
1. ESXi Hosts:
- Minimum: 3 hosts (for a functional vSAN cluster)
- Recommended: 4 hosts (for better fault tolerance and performance)
2. CPU:
- Minimum: 2 CPUs per host
- Recommended: More CPUs for higher performance, especially with multiple VMs
3. Memory:
- Minimum: 32 GB per host
- Recommended: 64 GB or more per host for better performance and to support larger environments
4. Storage:
- Cache Tier:
- Minimum: 1 SSD or NVMe drive per disk group
- Recommended Size: Approximately 10% of the total capacity tier. For example, if the capacity tier is 1 TB, the cache tier should be around 100 GB.
- Capacity Tier:
- Minimum: 1 SSD, NVMe, or HDD per disk group
- Recommended Size: Depends on storage needs, with consideration for future growth and data protection overheads.
Example Configuration
Host Configuration Example:
- CPU: 2 x Intel Xeon Silver 4210 (10 cores each)
- Memory: 128 GB DDR4
- Storage:
- Cache Tier: 1 x 400 GB NVMe SSD
- Capacity Tier: 2 x 2 TB SAS HDD
General Recommendations
- Network: Each host should have at least 10 Gbps network connectivity for vSAN traffic. Consider multiple NICs for redundancy.
- Storage Controllers: Ensure your storage controllers are on the VMware Compatibility Guide. Controllers should be configured in pass-through or RAID 0 mode.
- Disks: Use high-quality, enterprise-grade SSDs for the cache tier and a mix of SSDs or HDDs for the capacity tier, based on performance needs.
Storage Sizing Tips
- Cache Tier:
- Aim for SSDs with high write endurance since the cache tier handles most write operations.
- Size the cache to be at least 10% of the total capacity.
- Capacity Tier:
- Balance between SSDs and HDDs based on performance vs. cost considerations.
- Plan for data growth and overheads due to RAID configurations.
This overview should help we understand the hardware requirements for configuring a vSAN cluster.
Thank you
@@gnancloudgarage Thank You so much sir. 🙏🤝
Hi Sir. I want to deploy as a Stretched Cluster. I have four ESXi nested Hosts. Is it possible to create a stretched Cluster with nested ESXi hosts? Or else, three nested ones are enough to deploy a stretched cluster. Or three nodes are enough as we have resource constraints, I can't make more than four nodes for the Stretched cluster. One node is for site 1 and another node is for site 2, another node is for witness. Please suggest or guide me on deploying a stretched cluster with three or four nested ESXi hosts. thanks Sir. Please guide me in this regard.
Hi Sir,
Deploying a stretched cluster with nested ESXi hosts is indeed possible, though it comes with certain considerations and limitations. Let's break down your options based on your setup:
1. Stretched Cluster Basics:
- A stretched cluster typically spans two physical sites (Site 1 and Site 2) and includes a witness node to achieve quorum in case of a site failure.
- For VMware vSAN, which is often used in stretched cluster configurations, a witness node can be a lightweight virtual appliance or hosted externally.
2. Minimum Requirements:
- VMware recommends a minimum of three physical ESXi hosts per site for a stretched cluster. This translates to at least six physical hosts total (three per site).
- Each site should ideally have at least three nodes to maintain quorum and ensure fault tolerance during site failures.
3. Nested ESXi Host Considerations:
- Nested virtualization (running ESXi as a VM on another ESXi host) can be used for testing and some production scenarios, but it introduces additional complexities and performance overhead.
- Performance of nested ESXi hosts may not match that of physical hosts, especially in terms of disk and network I/O.
4. Your Specific Case:
- You have four nested ESXi hosts. To deploy a stretched cluster with these:
- Option 1: Consider deploying a 2+1 configuration (two nodes in one site, one node in another, plus witness):
- Site 1: 2 nested ESXi hosts
- Site 2: 1 nested ESXi host
- Witness: Can be a virtual appliance or an external witness node
- Option 2: If possible, expand to one additional nested ESXi host to create a more balanced 2+2 configuration:
- Site 1: 2 nested ESXi hosts
- Site 2: 2 nested ESXi hosts
- Witness: Virtual appliance or external witness node
5. Resource Constraints:
- Given your constraint of not exceeding four nodes, you'll need to carefully balance resource allocation and availability requirements.
- Ensure that each site has sufficient compute, memory, and storage capacity to handle the expected workload and provide redundancy.
6. Guidance:
- I recommend evaluating the performance implications of nested virtualization in your specific environment.
- Test failover scenarios and ensure that the chosen configuration meets your availability and performance needs.
Hi Sir. Would you explain What's a component node and witness node in the vSAN cluster?
Hi Sir,
Here’s an explanation of what a component node and a witness node are in the context of a VMware vSAN cluster:
Component Node
In the context of a vSAN cluster, "component" refers to the smallest unit of storage in vSAN, which can be a part of a virtual machine's (VM's) storage object. Each VM object, such as a VMDK (virtual disk), is divided into smaller pieces called components. These components are then distributed across multiple hosts in the vSAN cluster to ensure data redundancy and fault tolerance.
Key Points:
- Data Distribution: Components of a VM object are distributed across different nodes in the vSAN cluster to provide resilience against failures.
- Resilience: vSAN uses components to implement storage policies like mirroring, erasure coding, and RAID configurations.
- Storage Policy: The number of components and their distribution are determined by the vSAN storage policy applied to the VM object.
Witness Node
A witness node is a specialized component in a vSAN cluster that acts as a quorum to maintain data consistency and cluster integrity. It doesn't store actual VM data but holds metadata and acts as a tie-breaker to avoid split-brain scenarios.
Key Points:
- Metadata Storage: The witness node stores metadata about the vSAN cluster, such as information about the components and their states.
- Tie-Breaker Role: In a situation where network partitions occur, the witness node helps determine which partition should continue to operate to avoid data inconsistency.
- Deployment: A witness node can be deployed as a physical host, a virtual appliance, or a cloud-based service, depending on the vSAN configuration and requirements.
- Use Cases: Commonly used in stretched clusters, where vSAN nodes are distributed across geographically separate sites, or in 2-node clusters to provide quorum functionality.
Detailed Functionality
- Component Node:
- Each VM object is broken down into multiple components.
- Components are distributed across different physical hosts to ensure fault tolerance.
- If a host fails, vSAN can rebuild the components on other hosts using the available replicas or parity information.
- Witness Node:
- Essential for clusters that need a quorum, such as stretched clusters or 2-node clusters.
- Ensures that in the event of a network partition or host failure, the cluster can still maintain a majority and continue operating without risking data integrity.
- The witness node needs to be on a separate site or infrastructure to provide effective failure domain separation.
Example
- Stretched Cluster: Imagine a vSAN stretched cluster with nodes split across two sites (Site A and Site B). If Site A loses connectivity to Site B, the witness node (located at a third site or cloud) helps decide which site continues to be the active site. Without the witness node, both sites might incorrectly believe they should be the active site, leading to a split-brain scenario.
Conclusion
- Component Node: Fundamental storage units that make up VM objects and are distributed for redundancy.
- Witness Node: A special node that maintains cluster metadata and acts as a quorum to ensure cluster integrity and prevent split-brain scenarios.
Understanding the roles of component nodes and witness nodes is crucial for effectively managing a vSAN cluster and ensuring data resilience and integrity.
@@gnancloudgarage Thank you so much sir. 🤝🙏
Hi Sir. I created vSAN Cluster with nested 4 ESXi hosts like you did. However, I'm getting an error saying that Disks Groups are unhealthy. What would be reason? would you tell me Sir. Please Thanks.
Hi Sir,
Thank you for your question! If you're seeing an error that the Disk Groups are unhealthy in your vSAN Cluster with nested ESXi hosts, there could be several reasons for this issue.
Here are some common causes and troubleshooting steps:
1. Disk Compatibility: Ensure that the disks you are using are compatible with vSAN. Check the VMware Compatibility Guide to confirm this.
2. Disk Configuration: Verify that the disks are properly configured and recognized by the ESXi hosts. Each disk group should have one cache disk and one or more capacity disks.
3. Network Configuration: Ensure that the network configuration for vSAN traffic is correctly set up. All ESXi hosts in the vSAN cluster should have proper network connectivity.
4. Storage Policies: Check the storage policies applied to the vSAN cluster. Incorrect or conflicting storage policies can cause disk group health issues.
5. Health Check: Use the vSAN Health Service to run a health check. This tool can provide detailed information about what might be wrong with the disk groups.
6. Resource Allocation: Ensure that your nested ESXi hosts have sufficient resources (CPU, memory, and storage) allocated. Resource constraints can sometimes cause issues with vSAN.
7. Firmware and Drivers: Make sure that the firmware and drivers for your storage controllers are up to date. Incompatibilities can cause disk health issues.
8. Logs and Errors: Check the ESXi host logs and vSAN logs for any specific error messages that could provide more insight into the issue.
If you've checked all of these and still can't resolve the issue, you might want to look at the detailed logs
@@gnancloudgarage Thank You So much, sir. I really thank you for making an effort to answer my queries.
Hi sir. What will happen to Disk Group if the cache tier disk has been lost ?
If the cache tier disk in a VMware vSAN disk group is lost, the following events will typically occur:
1. Disk Group Degradation: The entire disk group will be marked as degraded because the cache tier disk is critical for the operation of the disk group. This will impact the performance and availability of the data stored in that disk group.
2. Component Rebuilding: vSAN will attempt to rebuild the components stored on the affected disk group elsewhere in the cluster if sufficient resources and redundancy are available. This process ensures that data remains protected according to the storage policy.
3. Data Accessibility: While the rebuilding process is ongoing, data availability might be affected depending on the redundancy and fault tolerance settings. In some cases, data might become inaccessible until the rebuild is complete.
4. Automatic Failover: If the cluster is configured with appropriate fault tolerance and enough resources, vSAN will use the mirrored copies of data to serve I/O requests, minimizing the impact on data availability.
5. Administrative Action Required: An administrator will need to replace the failed cache tier disk to restore the disk group to its optimal state. This typically involves physically replacing the faulty SSD and possibly reconfiguring the disk group in the vSAN configuration.
In summary, the loss of a cache tier disk in a vSAN disk group will degrade the disk group, trigger a rebuild process to maintain data protection, and require administrative intervention to replace the failed disk and restore full functionality.
@@gnancloudgarage Thanks Sir. 🙏🙏🤝🤝
@@gnancloudgarage Hi Anna. I'm interested to lean vSAN very deeply. would you teach me Anna. I'll pay the amount for learning vSAN. Please reply Anna. thanks
Looking for NSX and other VMware technologies like this
Hi Kareem MD,
Thank you for your interest.
Sure, will do that and in the meantime, refer to the subsequent sessions.
1. IT Infrastructure Evolution & vRealize Automation (vRA) 8.6.2 Architecture Overview
czcams.com/video/A2MoQ-UdXEA/video.html
2. Installing vRealize Automation (vRA) 8.6.2 with vRealize Easy Installer | Home Lab
czcams.com/video/jPz1e-H6nnw/video.html
1. Containers, Pods, Kubernetes, VMware Tanzu | Home Lab
czcams.com/video/fK75wijTPU8/video.html
Please do "View", “Like”, “Subscribe” to my channel and activate the “bell” notification so you don't miss new videos 😊
All the Best.
Hi. I have one doubt. what will happen to VMs and datastore if any one of Esxi hosts goes down or any outage occurs? Kindly answer to this query Sir. thanks
Hi Sir,
When an ESXi host experiences a failure or outage, the virtual machines (VMs) running on that host are affected. However, the impact depends on several factors:
1. High Availability (HA) Configuration: If High Availability is configured, VMs that were running on the failed host will be automatically restarted on other healthy hosts within the cluster. This process helps minimize downtime for the VMs.
2. Shared Storage (Datastore): Ideally, VMs should be stored on shared storage such as a SAN (Storage Area Network) or NAS (Network Attached Storage). In this setup, even if an ESXi host fails, the VMs' data remains accessible to other hosts in the cluster. The VMs can be quickly restarted on another host using this shared datastore.
3. Impact on Performance: During a failure, there might be a temporary performance impact on VMs as they are migrated to other hosts and restarted. This impact is usually minimal if the infrastructure is properly configured and sized.
4. Data Integrity: With shared storage, data integrity is maintained even if an ESXi host fails. This is because the VM's data is stored separately from the host itself.
5. Manual Intervention: In some cases, manual intervention might be required to recover VMs, especially if HA is not configured or if there are complications during the failover process.
In summary, in a well-configured VMware environment with HA and shared storage, the impact of an ESXi host failure on VMs is minimized, and VMs can be automatically restarted on other hosts without significant downtime.
Thank you
@@gnancloudgarage thank You so much sir. I have one more doubt, sir. Let's say we have three ESXi hosts in the vSAN cluster. Each ESXi host contributes 5GB (Capacity )to vSAN storage. There would be around 15GB. Data occupied around 13GB. In this case, if ESXi goes down, then what will happen to the data that the failed host contributed storage. What happens to 3GB of occupied data. Kindly answer to my question. Thanks Sir. Though the above-mentioned points are understood well to some extent. not fully.
@@jaganbj Sir
In a vSAN cluster, the data redundancy and availability are managed through storage policies that dictate how data is replicated across the ESXi hosts. When an ESXi host goes down, the impact on the data depends on the vSAN storage policies in use. Let me explain in detail:
### vSAN Storage Policies
The most common storage policy in vSAN is the Failure to Tolerate (FTT) policy, which specifies how many host or disk failures a cluster can tolerate. For instance:
- FTT=1: The data is replicated such that it can tolerate one host failure.
- FTT=2: The data is replicated such that it can tolerate two host failures, and so on.
### Example Scenario
- Cluster Configuration: 3 ESXi hosts, each contributing 5GB to the vSAN storage pool, making a total of 15GB.
- Used Storage: 13GB of data is stored on the vSAN datastore.
- Failure: One ESXi host fails.
### Impact of Host Failure
If the vSAN cluster is configured with an FTT=1 policy, it means that each piece of data has at least one additional copy stored on another host. Here's how this impacts the data:
1. Data Replication: With FTT=1, vSAN ensures that there are at least two copies of each piece of data across different hosts. Thus, for the 13GB of data, there are replicas spread across the remaining two hosts.
2. Host Failure:
- When one host (e.g., contributing 5GB of storage) goes down, the cluster loses access to that host's data temporarily.
- However, because of the FTT=1 policy, the data is still available through the replicas stored on the other two hosts.
- The 3GB of data that was on the failed host has replicas on the other hosts, ensuring no data loss.
3. Rebuilding Data:
- vSAN will automatically start rebuilding the missing replicas on the remaining healthy hosts to maintain the FTT=1 policy.
- This means that the cluster will redistribute the data to ensure that the redundancy is restored, provided there is enough free capacity in the remaining hosts.
4. Performance Impact:
- During the rebuild process, there might be a temporary performance impact as the cluster works to reestablish redundancy.
- If the cluster does not have sufficient free space to rebuild the data, it may not be able to maintain the specified FTT level until the failed host is brought back online or additional capacity is added.
### Summary
In your scenario:
- With an FTT=1 policy, the 3GB of data on the failed ESXi host will still be available through its replicas on the other two hosts.
- vSAN will work to rebuild the lost replicas on the remaining hosts, ensuring continued data availability and redundancy.
If a different FTT policy or other configurations (such as RAID settings) are in use, the specifics might vary, but the general principle of data redundancy through replication holds true in vSAN.
Always ensure that your vSAN cluster has adequate free capacity and is configured with appropriate storage policies to handle host failures without data loss.
@@gnancloudgarage Thank you So much sir. 🤝🙏🙏
@@jaganbj Most welcome Sir 🙏🤝
Hi , how can we join the session live with you 1 -1 ??
Hi,
Thank you for your interest in joining a live session with me!
At the moment, I don't have any one-on-one sessions scheduled, but I appreciate your request.
@@gnancloudgarage let me know once you are available for team session as well , i need it for tanzu kubernetes etc...
excellent teaching
Thank you so much Kareem MD :-)