Weathering the Storm: Using HPE SimpliVity for Data Efficiency and Disaster Recovery
When most people picture living life on a ranch, some iconic images come to mind. But agriculture, cowboys, and sprawling acreage just scratch the surface of what life can be like, and the different types of work invested in keeping such a wholesome, simple way of life. Unless you’re familiar with Schroeder-Manatee Ranch (SMR), you may not realize just how much technology and IT know-how goes into maintaining our many assets across multiple industries.
As a mostly self-taught IT professional, I achieved certification over the past 15 years and worked my way up to become SMR’s systems administrator. I’m challenged every day, because there’s always something new to learn. For example, up until a few years ago, I had never even heard of the term hyperconvergence. Today, I’d point to it as one of the biggest positive shifts we’ve made. Hyperconvergence was totally a new concept—a solution to my need to manage data as our older system started to show signs that it could no longer keep up with our needs.
Problems Piling Up
Our IT department has always been very efficient, streamlined, and had just what we needed to get the job done. We have a little over 200 users, who are managed by our three-person IT team. All employees are connected on one big network through site-to-site VPN tunnels. We manage servers, emails, and services like CRM and SharePoint. We're responsible for all the copiers, cell phones, and printers as well.
We had a setup where our five blade servers—organized in a cluster—were all running Microsoft's Hyper-V virtual machines (VMs) and we started to get low on resources with those five blades. At maximum, we could only comfortably run 10–12 VMs concurrently. The blade enclosure had spots for more blades themselves, so we initially thought that we could just throw more hardware at the problem. By then, we anticipated that we'd have more resources to add more VMs.
The other issue we had was storage in general. At the time, the IBM blades had hard drives just big enough for the OS. We had a separate SAN solution—which was even older than the IBM blades—and only had about two terabytes of space total. It was already getting close to being maxed out. I cringed as I watched the SAN storage slowly dwindle, which housed our critical files such as email databases, SQL databases and the main file server for our headquarters office.
The system also lacked a built-in backup solution, so we relied on third-party backup solutions. Since day one, that had always been frustrating. Even when we switched to a cloud-based solution and tried to get away from on-site tape/disk backup solutions, the backups would just run all night and into the next day. I’d come in every morning and have to check which backups were still running and which ones had failed (which seemed to happen fairly often for one reason or another). Disaster recovery (DR) weighed heavy on my mind. It troubled me to know that we weren’t prepared for the worst.
When Throwing Hardware at the Problem Won’t Work
With all of these issues in hand, we went to our hardware vendor, told them about the problems we faced, and asked for a quote for more blades and a larger SAN. Our focus was getting our hands on more RAM, CPUs, and a new storage solution because we’d dealt with the same system longer than we should have.
Luckily, we’ve worked with this vendor for a while. They know our business and about the potential solutions that exist in the market. They were able to challenge our assumption that more hardware was the solution. We were recommended SimpliVity’s (now HPE SimpliVity) hyperconverged infrastructure solution. Our vendor scheduled the initial meeting and demo with the SimpliVity engineer and salesperson, and they came to SMR.
We’d never used a product like this before, but right away the concept of hyperconvergence sounded interesting. I liked the idea of consolidating our setup. It would be one unit to house everything, including the storage, memory, and resources.
Luckily, we were still a smaller company, so we didn't have to go through an extensive competitor and cost analysis. The executive staff trusted our judgment and, since it fit within our budget and hit our key requirements, they let us go ahead and get our first pair of SimpliVity systems so we could try them out.
The Simpler Life
The initial roll-out went smoothly, with HPE SimpliVity’s engineers coming out to SMR to help install all the necessary hardware and configuration. They also helped me create what essentially became our template VM. We created a VM, installed Windows Server, installed all necessary updates and third party software, then saved it as a template. Now, whenever I want to get a server up and running, I can use this template and have everything done in 10–15 minutes.
The template only needs to be updated a few times per year when I run a Windows update. To put it in perspective, every time I created a new VM in the past, I would have to run through their little wizard, then connect to an ISO file, install the OS from scratch, and do all the updates. It was incredibly time consuming for what should have been a simple task.
Once we had installed the hardware and created the template, our next step was to convert our existing servers to the new system. The Exchange server couldn’t handle a standard migration, so we had to set up a new one in order to migrate the mailboxes accordingly. HPE SimpliVity’s engineers walked me through the process so I could see how it was done. Then, I continued along myself as my schedule permitted. We tried to do everything at night, just to minimize the impact on the users.
The very first thing I noticed once everything was installed was that the management console was much more user friendly and convenient with all I needed built into one screen. I managed storage, my VMs, and my backups from one console. This new setup made the entire process streamlined and more efficient compared to our last system where I was dealing with different consoles and additional software to manage one system.
The servers were more responsive in the new system, and being able to configure the backups and store them right inside the system itself was exactly what I was looking for. Also, the learning curve for the vCenter software was very minimal. It's user-friendly and HPE SimpliVity also offered some initial training, so I was able to jump right in.
We now use three pairs of HPE SimpliVity systems, which are all managed in the same console, and the most impressive improvements we’ve experienced so far are our data efficiency and data deduplication ratios. Each pair has saved a lot of storage—up to 632 TB of data. Their current deduplication and efficiency rates are:
1. 64.3:1 deduplication / 105.5:1 efficiency
2. 76.2:1 deduplication / 99.9:1 efficiency
3. 56.7:1 deduplication / 79.2:1 efficiency
We’ve achieved some outstanding numbers for the amount of deduplication we have. And it’s proof that we were on the right track in terms of cost savings, storage, backups, and disaster recovery.
Weathering the Storm
Of everything we’ve been able to accomplish, disaster recovery is maybe the most tangible. With our former backup data protection software, there were always problems. It was temperamental and had us clinging to old media that was inefficient and a pain to take offsite. We now back up our VMs to another site on our network, which also has a pair of HPE SimpliVity systems. It was initially bought, and set up, as a disaster recovery site. With the communication between the units, it’s now only backing up any incremental changes; the critical VMs are backed up every 10 minutes. Now, it's so efficient to both backup and restore something quickly.
The current DR setup is a relief because I never have an issue with the backups now. I simply check them every week just to make sure everything still stays protected. I've never had issues with backups running or with storing data. That alone is a huge weight off my shoulders.
With the HPE SimpliVity solution, we have no issues trying to restore anything we need, including individual files. This functionality couldn’t have existed at a more crucial time than when we got hit with Hurricane Irma in 2017.
Since all of our sites are within a few miles of each other, we knew we were going to feel some kind of impact from the storm, so we got prepared ahead of time. We already had our backups of our critical VMs backing up every 10 minutes over to another facility. This other site was the newest, and if anything could withstand the hurricane and survive a power outage, it was this one.
When the storm came through, we shut down the physical servers we had at each site and monitored the power. Eventually, the power went out. At headquarters, we have a little bit of lead time before the UPS loses all battery power, so I logged in when I realized the power wasn't coming back on. I shut down all the VMs and physical servers and did one last final backup over to this remote site that has a natural gas generator to keep the equipment up and running indefinitely. When a manual backup is done, it doesn't ever go away so I knew it would be safe for days, weeks, or more. We can be down for a day or two before we start to panic, so we kept our fingers crossed that the power would come back.
By the end of the second day, we realized that wasn't going happen and decided to start up the server at the remote site that still had power from the generator. Normally, if it was just a regular file server, it would have been easy to just fire the server up, but because it's Exchange, there's a lot more involved such as configuring firewalls, updating DNS and MX records, etc.
Our initial concern was just getting the Exchange server powered up first and then worry about our other critical VMs after. Since we already had a backup onsite, we just restored the VM, pointed it to the new data center, and it was up and running quickly. We were at the verge of powering on more VMs when the power finally came back at our HQ office.
The alternative, with our old system, would have been a disaster. We had no redundant hardware for our old blade system and even if we did, we couldn’t possibly have moved as quickly and recovered our systems as fast as we did with HPE SimpliVity. The ability to bring up those servers so quickly proved to the executive staff the power of our new solution. Now, we’re even more confident and prepared for the next time something like this happens.
Less Stress and More ROI
When I look back at the migration to HPE SimpliVity and hyperconverance, there are three key lessons we’ve learned since adopting HPE SimpliVity.
First, plan for more than you think you’ll need. Storage eventually becomes a concern because you don't have unlimited storage on these units. So buying more storage initially would've been a little more helpful. We initially only thought we would run our critical VMs on the HPE SimpliVity systems, but after seeing how much better the HPE SimpliVity solution was we started moving more and more VMs over to them as well.
Secondly, if you work with the HPE SimpliVity engineers, listen to what they have to say, take their recommendations, and it should be a very smooth process. We trusted their best practices, and that’s part of what made the process so easy.
Finally, if you have questions, don’t be shy about taking advantage of any training or customer support HPE SimpliVity has to offer. The support was very knowledgeable and hands-on, and the basic training set us up with a solid foundation.
I see no real downsides to hyperconvergence. It makes so much more sense than any other option we considered. We’re seeing a meaningful ROI, while at the same time improving the experience for our end users. Being in IT, that’s what it’s all about. My job is to make their jobs easier. And as SMR operations continue to grow, we’re excited to finally have an IT and virtualization environment that can keep our team as nimble as ever.