The storage policies allow the administrator to define which VMs have specific storage requirements. For example: Our critical VMs have an increased flash read cache percentage enabled. This improves the overall performance of these machines. The ability to specify policies for every kind of VM in your data center improves storage efficiency, as well as improving performance, redundancy, and so on for specific VMs. With traditional SANs, configuring this was only possible on a LUN level. With vSAN, we can do this on the VM objects themselves.
One of the things that surprised me was the way vSAN handles a disk failure. It auto-rebuilds the vSAN objects when a failure has been detected. (Note: There are two kinds of failures, and this has a different effect on the rebuild timer.) But, in the end, the cluster is self-healing without any user input needed. The only thing that is affected is purely the raw storage that is lost with the drive.
Improvements to My Organization
The ease of managing and configuring vSAN. This means that all our VMware administrators are now able to do the daily maintenance and operations. Previously, only a couple of IT administrators were responsible for maintaining our previous storage solution and the complex tasks that came with it.
Room for Improvement
- The daily maintenance can be high, especially due to the lack of documentation and reporting in vCenter, and only on the vSAN health page.
- If the vSAN cluster can’t self-heal due to an internal error, we can’t repair the vSAN cluster ourselves.
- A case with VMware is always needed to fix the issue, resulting in an increased time to resolve. This can be very time-consuming.
- I would like to see more documentation on the errors, impact, and solutions. This could improve the product knowledge.
- Some essential storage features (deduplication/compression) are only available on all-flash vSAN clusters. These limitations need to be taken into account when sizing and designing your environment.
Use of Solution
We have been using this solution for a year.
Keep a close eye on the vSAN HCL. As vSAN is continuously in development, the HCL changes as well and so the HCL gets updates.
When you are planning to upgrade the vSAN version, all other components (ESX version, server firmware, server BIOS) need to be checked to see if they are all on that version’s HCL.
Scalability on vSAN is extremely easy. If the host is compliant with the prerequisites (one SSD and one spinning disk), it will be accepted by the cluster instantaneously. All raw storage will be committed to the vSAN data store and directly available for usage.
In terms of sizing the cluster, as deduplication and compression are only available on all-flash arrays, this can heavily impact the storage capacity of the vSAN cluster.
Since we chose a hybrid-configuration, the lack of deduplication and compression caused a storage growth that exceeded the limits quite rapidly. We had to scale up and address the issue in other ways.
Customer Service and Technical Support
Technical support is good. When encountering issues with vSAN, 99% of the time a VMware support case needs to be opened. All of the standard steps of a support case are run through. In the end, a VMware engineer will solve the issue with you and bring the cluster back to a fully healthy state.
Our previous hyper-converged system broke down due to a power failure. A new system was needed. vSAN was the logical choice, as we are a VMware Partner.
The way VMware integrated the vSAN hyper-converged storage functionalities in their vSphere Kernel is really revolutionary.
It allows the environment to scale out on storage resources when the business needs it. You no longer have to buy those expensive traditional SAN setups scaled for the “future requirements” that you had in mind at the time.
Even an IT administrator with some basic VMware experience would be able to set up vSAN in just a couple of minutes. This is one of the easiest setups I have had in a while.
Other Solutions Considered
We had previous solutions, but vSAN was the logical choice.
I would definitely recommend vSAN to others. The old, cumbersome, and traditional storage environments are done and belong to the past. Hyper-converged is the next big thing. It is more cost effective, easier to manage, and scaling up can be done almost on the fly.
I recommend going for an all-flash vSAN setup, if the budget allows it. Some vSAN features like deduplication/compression are only available on an all-flash configuration.
With the falling GB/$, an all-flash is becoming the evident choice. The benefits are there (more features and all-flash performance for all VMs).
Disclosure: My company has a business relationship with this vendor other than being a customer: The company is a VMware Enterprise Solutions Provider Partner.
Feb 16 2017