Posted On May 28, 2024

VMWare Deployment Overview

kimconnect 0 comments
blog.KimConnect.com >> Virtualization >> VMWare Deployment Overview

1. Configure VMWare ESXi
– If dedicated storage card(s) is/are available, configure LACP/PAGP etherchannel or bond on the attached switch(es) for active-active (not active/standby due to probability of non automatic activation or when standby connection(s) has silently detached) in failover mode.
– Configure network switches to seggregate these VLANS: VM (dedicated to guest virtual machines), Management (can be on the same subnet as the VM network), iSCSI (or SMB/NFS storage subnet with recommended 9000 BTU for enhanced performance), vMotion (for guests migration), heartbeat (WMWare’s “Active-Passive” architecture of high availability)
– Install 10G/25G/32G/ PCIe SFP+ card(s) into server motherboard. Best practice is to have redundancy of multiple NIC’s bound into a vSwitch.
– Update system BIOS
– Download and install ESXi 8.0 onto hardware and active license
– Install updates to bring kernel security to compliant levels
– enable SSH, assign Name & static IPv4 on Mgmt Interface

2. Integrate ESXi into Active Directory
– Create a new group in Active Directory as ‘ESXi Admins’ (or someting sematically similar), and add users into that group
– Join VMWare host(s) into Active Directory

3. Install vCenter & vSphere Client
– Download & Install vCenter ISO’s from the vendor & add licences
– Patch vCenter
– Add Host(s) into Datacenter group(s)
– Create 2 new groups in Active directory as vCenter Admins and vCenter Read-Only Users. Populate these groups appropriately
– Join vCenter onto Active Directory
– Create 2 new groups in vCenter to include the 2 Active Directory groups generated previously

4. VM Storage
– Internet Small Computer System Interface or iSCSI server is the common method to present Logical Unit Numbers (LUN) and ‘targets’ to guest VMs due to plug-and-play features. This can be done on the storage appliance (HPe Nimble, NetApps, Windows Server, etc.). Multiple LUNs can be combined into a single target for the clients to attach.
– Best practice is to dedicate at least 2 physical NIC’s into the iSCSI network (vSwitch). Hint of this setup is to create each vNIC being associated as 1:1 with a physical NIC, then several dedicated vNICs can be added into the iSCSI vSwitch. Dual fabric design is the standard.
– It is recommended that large LUN’s of 16TB+ are presented as targets to be connected from the ESXi hosts and controlled by vCenter storage cluster. This simplifies management as compared to running initiators from guest VM’s.
– Affinity and anti-affinity rules must also be setup for security & I/O performance purposes. A single LUN should be localized on one controller, instead of multiple controllers as that would lead to higher blocks processing overhead (lower I/O results).
– Certain deployments would include NFS, SMB, or vSAN. those are pieces of cakes which may not merit detailed scripts in this overview.
 – Benchmark using vRealize Operations Manager or vSphere Storage Performance Monitoring to determine when to rebalance LUN’s on underutilized controllers
– Label LUNs basing on I/O ‘class’ or speed, redundancy, and backup schedules. This naming convention would assist in deciding which blob to use for database, general compute, or low-tiered machines.


5. vSphere Site Recovery Manager
– Group VM Clients into folders
– Config Replication for guest VMs
– Setup Protection Groups with replication type “vSphere Replication (VR)”
– Create Recovery Plan & associate the Protection Group with it. Then, configure each guest VM’s with destination IP’s, storage, and priority levels

Leave a Reply

Your email address will not be published. Required fields are marked *

Related Post

Configure VPN from AWS to CPE (Juniper Router)

1. Create VPN connection from AWS portal and download the configuration file2. Import the configuration…

How to Upgrade Kubernetes Ingress Nginx Deployed via Helm

# How to upgrade ingress-nginx: helm upgrade --reuse-values ingress-nginx ingress-nginx/ingress-nginx -n ingress-nginx # Sample output of a failure scenario: brucelee@k8-controller:~$ helm…

Kubernetes: Cert-Manager x509 ECDSA verification failure

Symptoms Error from server (InternalError): error when creating "kimconnect-cert.yaml": Internal error occurred: failed calling webhook…