What is our primary use case?
I am working for a public cloud provider and am supporting their infrastructure. The company's cloud is deployed on VMware products. Essentially, it is VMware virtualization infrastructure and they are selling public cloud space.
Customers use the service to have access to a public cloud that is local, in their country. If, for example, they don't want to use AWS or Azure, then they can opt to use this service. In return, they have full control of their data and infrastructure.
We use several products in the VMware suite including ESXi, vCloud Director, NSX, vRealize Operations Manager, vRealize Operation Log Insight, vRealize For Business, and vSAN. The company runs 80% of VMware products.
How has it helped my organization?
This solution has helped improve our organization because we are a successful local cloud provider and the number of customers that have joined our cloud is increasing. Our customers know that we're running on VMware and we haven't faced any issues yet. Moreover, most of our customers' businesses are doing well. Overall, we have done well with VMware.
We do our daily proactive monitoring using vRealize Operations Manager. It provides us full insight in terms of what is happening in our operations, including the details contained in the logs. VMware vCenter also helps us with proactive monitoring.
Proactive monitoring has helped us to avoid downtime, especially because we follow the best practices described by VMware. When you follow best practices, you won't face many problems. The overall downtime depends more on your support and handling of the product, rather than the software. We are running on a cluster to help avoid downtime.
vROps has helped us to place workloads efficiently, although our users do not have very large workloads. We are running two environments and we are able to handle the users and workloads that we have.
Using vROps has helped to increase our VM density within our clusters. VMware provides a solution where you can create a cluster, whether for storage or compute, and nodes within the cluster are monitored. If there is a node that goes down then it is automatically kicked out of the cluster. Before the host goes down, vRealize creates a replacement. It has three copies of each disk in different host nodes, and it will automatically trigger one of the copies. This makes sure that the system is fault-tolerant and the VMs won't have any problems. Also, if the Distributed Resource Scheduler (DRS) is turned on, the VMs will balance the resources and try to avoid downtime.
We have been able to replace other apps with vROps because it is, pretty much, doing everything. For example, we had another monitoring tool that was running on each of our nodes. Its job was to trigger alerts and display information. vROps replaced it and is even more powerful. If the network goes down between our office and one of our data centers then it is able to detect that. It will provide you ways to get a deeper understanding of the issues, and it will suggest resolutions. There are other products that vROps has replaced, including solutions for resource planning and load management.
All of our products are integrated with vRealize Log Insight. It integrates with the other components including vCenter and NSX and retrieves all of the logs. From there, logs are stored in the system and when you have problems, you can deep dive and perform a log analysis. You just have to know the keywords you are looking for, which components, and the hostname, or the host IP address. It will report all of the information in the log that is related to it.
Troubleshooting works well with vRealize Log Insight, provided that all of the component drivers are updated and the service packs are all installed and running. When we configure the integration, we have to verify where the logs are coming from. As long as it is set up correctly, troubleshooting will not be a problem.
What is most valuable?
There are four main components that we use. The first is the hypervisor, EXSi. It is the most important part because this is the virtualization medium. Without it, you cannot set up or deploy your virtualization environment.
The next component is NSX, which allows network virtualizations to provide your tenants with the ability to manage their own network.
We have vSAN for storage virtualization, to create clusters.
We also have a tenant portal, vCloud Director, for self-management, including payment. Tenants are able to control and manage their virtual data center by themselves without the involvement of the service provider.
What needs improvement?
We have faced one problem when integrating with vRealize Log Insight, where the logs are not collected because the component drivers are not updated. Rather than give us the updated logs, the old ones are retrieved. The integration with vRealize should be more seamless.
One area that needs to be improved is vCloud Director, as it has very weird behavior sometimes. All of the other components are stable and you can predict their behavior. However, with vCloud Director, you can't always predict what it's going to do. For example, there are times when we thought that it was collecting information about the network, compute, and storage resources from vCenter, as well as information about the nodes, but it doesn't always work as expected.
The last time we had a problem with vCloud Director, we were unable to get the snapshot of the VM. From the backend, everything appeared to be running fine. This is an instance when we had to contact VMware support in our time zone, and they were able to help us.
You can find information about some of the problems with vCloud Director in the Knowledge Base articles that include various workarounds. VMware advises that when you face these kinds of problems, contact them to raise a ticket and they will come and fix it. The component is very sensitive.
In the past, when we have raised priority one tickets and they have sent us level one engineers. This wasted time because the L1 was only able to perform the troubleshooting steps that we had already completed.
For how long have I used the solution?
I have been using VMware for two years.
What do I think about the stability of the solution?
The agility of the software components is fine and there are no complaints so far from us, so far. In general, the stability is good.
The problem that we faced recently was that we were running out of time for support. We were still running on ESXi 6.5, and support was ending at the beginning of this year. We had to upgrade our infrastructure and we had our hands tied. We could not move forward until the upgrade was complete, so it was a marathon of activity. This included adding two different sites and it required that all of our regular activities were interrupted. Ultimately, however, it worked and everything is now good.
Sometimes, there are issues with stability that arise from the hardware. We are located in Kuala Lumpur and our data center is based in Bangkok, Thailand. Although it has been okay, we have encountered a few power interruptions. We are using HP machines, which are good, but there have been troubles with some of the SSDs. When that happens, because I install the operating systems using a USB, sometimes the drive is misplaced. These are the types of issues that we face more often.
In the case of any downtime with a node, the data center operator is there to quickly overcome and resolve the issue. Once we realize that a node is down, a replacement is automatically started and communicates with the other hosts. This allows us to avoid interruptions in the operation and in the business. Once things are repaired, and the original node is put back into the cluster, everything goes back to normal.
What do I think about the scalability of the solution?
So far, the product has been good and we haven't faced much in terms of complications. Our environment is new and we don't have millions of users, yet. We are still growing. I'm not sure if we'll face multiple problems when we reach one million users or even 500,000 users, but so far, everything is okay. We've managed to handle the workloads, and we've managed to satisfy our customers.
We have more than 100 customers using it and at this point, everything is running smoothly and the number of workloads is okay for our resources. As more customers come, we will increase our resources and expand our usage.
Overall, scalability is very good and it's one of the reasons that I like VMware products.
How are customer service and technical support?
Our experience has been okay because we have received support for any problems that we have had. Also, we were able to get support from anywhere. It is not only available in our time zone, but we can get support from elsewhere if, for example, we need it overnight. Global support is available from anywhere in the world.
I can say that we have had a few bad experiences, but overall, you cannot take two out of 100 and say it's bad. On the contrary, overall it is good.
I don't know how it works in other time zones, but our time zone is supported by India. I have found that sometimes, you have to push them hard. For example, we have raised a P1 ticket and in response, they sent us an L1 engineer. When a ticket is priority one, it means that the situation is critical and the business is impacted. If you send a Level One engineer in a case like that, it will waste time because they will perform the troubleshooting steps that we have already completed. This has happened to us a couple of times.
Which solution did I use previously and why did I switch?
VMware is the first solution that I used for virtualization.
How was the initial setup?
VMware has introduced the Server Appliance, which allows you to deploy very quickly. We just need to import the appliance VM and deploy it. Traditionally, we had to create a Windows machine, and there were several things to configure, but they now have their own operating system called Photon OS. It shortens the length of time required for deployment.
The initial setup of vCloud Director is a bit complex. Sometimes when we have a problem with it, we can't fix it. VMware themselves suggest raising a ticket when an issue arises, and they will come in to fix it.
When we first implemented this product, we came up with a plan and submitted it to VMware. The VMware team reviewed it and advised us of the best practices. We developed a set of instructions that includes deployment and updating the solution and re-submitted it to them for review. It was finalized and we follow this plan whenever we deploy it. Whenever we encounter problems, we raise a priority one ticket and they come to help us with the problem.
What about the implementation team?
The first time we deployed this solution, the local VMware team assisted us.
What was our ROI?
This service has been running for approximately four years and they are making a profit. Otherwise, they would discontinue it. They are planning to expand so there has been a return on investment, although I don't have the exact figures.
Which other solutions did I evaluate?
We did not evaluate other options before choosing VMware.
What other advice do I have?
We offer a variety of services for our customers including Kubernetes monitoring and management. However, at this time, we do not have any customers who opt for it. What we provide depends on the customer's requirements. If they want to include VMware with their machines, we deploy the tenants. We promote all of the products, including that for Kubernetes monitoring and management, but nobody has yet requested Kubernetes. I expect that because we are promoting them, our users will understand the utility and plan to use them in the future.
VMware updates their product every one or two years, and I think that they are ahead of us in terms of what features are needed. Overall, I think that the product is very good. In the future, we'll have experience with the functionality of all of the new features that VMware is coming out with.
The biggest lesson that I have learned from using this product is that if you want to have a private cloud, VMware is the best option. It is the most stable and the best choice for a private cloud investment.
I am planning to open my own cloud in my country, which will help the local community because many government agencies will not use the public cloud. For this, I'm thinking that I will be using VMware.
I would rate this solution a ten out of ten.
Which deployment model are you using for this solution?
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?