What is our primary use case?
I am dedicated to Cisco and work with multiple products as a consultant from one of the managed service providers.
Mostly it is compute, so I take care of this part only.
The performance issues concern site and IOPS requirements, in the case of PCI cards to the VMs. Sometimes expected IOPS are not there, and storage IOPS are not present. It is not very frequent, but sometimes we encounter disk failure issues with the Dell vSAN Ready Nodes, and it generally takes some time to change the disks.
My job is to be a solution architect, proposing and building items. It is more about designing solutions and implementation, not too much in daily operations and troubleshooting.
I deal with multiple environments. Sometimes we see 20 plus nodes around 32 or even more. Sometimes it is a small cluster, two node cluster, or three node cluster. I would say all kinds of environments based on customer needs.
Regarding existing environments and third-party solutions for backup and monitoring, primarily we do the installation of the VMware platform, VMware vSphere ESXi. VMware has one of the largest ecosystems with on-premises or private cloud providers. Integration perspectives require some API-based work.
It is a starting story for AI, but we are taking some use cases leveraging the Dell vSAN Ready Nodes along with GPU cards from NVIDIA. We are exploring other providers such as AMD, and Intel is also doing enhancements based on AI. We are recommending customers leverage nodes which have AI-supported applications using Dell vSAN Ready Nodes that support AI-based workload primarily. The VMware container technology integration allows VMware customers to enable AI-based applications, but that is with the latest generation only, not with older ones.
What is most valuable?
First is the administration which makes it easy compared to using third-party storage. Second is the storage policies, with different flavors available all within a single management console when we want to configure different sets of policies as per workload requirements. I would say in-house management, administration, and storage policies are the three key differentiators compared to normal non-Dell vSAN Ready Nodes where we are leveraging external storage.
What needs improvement?
There are no specific issues, but in some cases we have seen hardware failures and performance issues where performance suddenly deteriorates.
For how long have I used the solution?
What do I think about the stability of the solution?
Sometimes we experience performance issues and disk failure issues frequently. Sometimes it impacts another disk group situation. For example, a disk group for five disks can be problematic, and customers need to tolerate downtime if the disk is not available, leading to space issues.
What do I think about the scalability of the solution?
Scalability involves two types of environments: VMware Cloud Foundation (VCF) and non-VCF environment where we use VMware Validated Foundation. With VCF we have significant automation built-in within the VCF console within SDDC manager as the primary management console. For non-VCF environments, it depends on customer choice. For small environments, we do manual scalability and sometimes use Terraform scripts to keep the host ready and add in the cluster automatically. For large environments with frequent scalability actions, we propose the Terraform solution that builds ESXi, does necessary configuration, and automatically starts up for associating with the existing environment. If scaling is infrequent, with one or two hosts per month, it is a manual job. From a challenge perspective, it seems easier, with just a few pre-checks regarding make and model compatibility, disks, and server health on the hardware level.
How are customer service and support?
Not much interaction, as we work with multiple vendors including Dell, HP, Cisco, and Lenovo.
How would you rate customer service and support?
What was our ROI?
The cost is higher compared to non-Dell vSAN Ready Nodes, but when compared with external storage solutions, there are significant benefits. For example, if a customer needs 1,000 TB of storage or 1 petabyte as an external storage solution, they would need additional storage fabric, boxes, and controllers. With Dell vSAN Ready Nodes, compared to two-tier architecture, customers get 30 to 40% savings. This applies not only to CAPEX but also to operational costs over 5 years. The OPEX benefits are around 30 to 40% because the same engineer who manages vSphere can handle Dell vSAN Ready Nodes, eliminating the need for additional support engineers.
Which other solutions did I evaluate?
We work with HP and Dell, and sometimes it is a customer decision, while other times we have to recommend based on customer budget and terms with respective vendors.
What other advice do I have?
HCL Tech has now become senior in the field.
My recommendation would be especially for private and public cloud vendors, as Dell has their own AI solutions. I would recommend customers looking for AI-based workload to leverage and experience Dell vSAN Ready Nodes. They should start with a smaller scale then scale up based on their comfort and experience.
I rate Dell vSAN Ready Nodes 8 out of 10.
My overall rating for Dell vSAN Ready Nodes is 8.5, considering all perspectives.