What is most valuable?
The cover points feature in VxRail is remarkable. It's unique. It has an intervention failover system as well as an automatic failover system, reaching clusters existing in VxRail. This makes all files act as a single file with a large and huge resource, and, of course, with customized administration, configuration, and resources. It provides automatic failover for redundancy and data recovery.
What needs improvement?
If they can provide deduplication compression through the traditional hard drives, as Cisco does in the HyperFlex system, it will be very cost-effective, especially when it comes to archiving workload.
VxRail doesn't allow the mixing of old flash clusters and hyper clusters. When I'm starting with an old flash cluster and it comes to archiving workload, I will also need to attend the new cluster. So, I either manage two different clusters, or I pay and upload my work with the archiving mobile hard drive, which is not cost-effective at all. The main key is to allow mixing between two types of clustering, like Nutanix, or allow deduplication of completion over the period of shared hard drive on SAV. It will be much better.
For how long have I used the solution?
I have been using this solution for four to five years. I have used three generations of Dell servers.
What do I think about the stability of the solution?
It is very stable and durable. It only depends on vSAN, which is the number one software that defines storage.
What do I think about the scalability of the solution?
Currently, more than 500 users are using VxRail in our company. It's capable of delivering for all types of workloads. Technically, it depends on the hyper-converged instructions. This means that you have 100% assurance of its compatibility with all of its components. It should also carry all types of workload dispersing, that is, from the normal traditional virtual machines to high-performance computing, such as HEP workload, heavy database, artificial intelligence, and business analytics.
How are customer service and technical support?
They provide good support. You can reach them, especially if your system is at ESRS, EMC functional support. You can just chat with one of their technicians. They collect the logs and discover the issue. It takes almost a couple of hours from opening the ticket to resolve it. They are very good.
The hardware replacement takes 24 hours. They have their own stock here in Egypt.
How was the initial setup?
It is easy to install and implement the VxRail clusters. The initial setup was a piece of cake for us.
What about the implementation team?
We manage the storage, compute, and virtual machines as well as networking through the perfect channel.
We do all kinds of deployments. We check whether the customer wants to deploy it on-premises, cloud, or integrate with the public cloud to tier and replicate.
What's my experience with pricing, setup cost, and licensing?
VxRail is very cost-effective and affordable in the long term. It is more recommended when it comes to financial life, but it may depend upon the license. VxRail comes with VMware licensing, which may not be that cost-effective as compared to others. With VMware, it's an auto check competition. VMware is an expensive solution, especially for Nutanix. Nutanix have their own hypervisor called Acronis, which is very cost-effective against the VMware.
Nutanix is cheaper for the hardware but not for the software. If you ask the Nutanix partners to deploy Nutanix over Cisco servers or Dell EMC servers, the cost will be higher. Nutanix wants to compete financially. Therefore, they propose their software over the Supermicro server, which is a very cheap Chinese server. In fact, I don't like their terms of service.
HyperFlex has the highest price, and it is very expensive. I don't know why. It may be because this is a UCS system, which comes from Cisco and is already expensive. When it comes to HyperFlex, they need the labor to deploy Hyper-V, Citrix, or any other hypervisor.
Which other solutions did I evaluate?
I also deal with HyperFlex and Nutanix. In fact, I prefer VxRail.
When comparing with HyperFlex, VxRail is much alike HyperFlex. It is very cost-effective, and it doesn't have too many conditions like HyperFlex. When you start with building clusters in HyperFlex, you stick to the selected nodes forever. It is not the same in VxRail. You start with pNode in VxRail, and then you add eNode, sNode, etc.
HyperFlex has its own limitations. They say we can create up to 64 nodes, but, in fact, there are only 52 storage and 52 nodes compute with no mixing between two workloads. On the other hand, in VxRail, you can really create up to 64 nodes, which means the double amount of nodes to carry more servers, more computing in the clusters.
There are too many concerns about HyperFlex, especially related to performance. HyperFlex source the deduplication compression. You don't have the option to enable or disable the deduplication compression, which means that deduplication ends the performance. In VxRail, you can enable or disable the deduplication compression. So, you can gain a net performance against the storage, and you can move the storage against the performance. You can balance the full configuration.
When it comes to the software, Nutanix is great. The main concern is that Nutanix doesn't have its own hardware, and it is integrated with different types of servers to deploy its own system. Nutanix just has a contract with Noble, Supermicro, or HP to develop its own system, which is okay for some types of users. However, many types of users request and prefer the full software or hardware that comes from a single vendor so that they can achieve the maximum and ultimate support.
What other advice do I have?
I would rate VxRail an eight out of ten. They should allow the deduplication compression over the hard drives and mixing of the hyper and the old flash clusters.