What is our primary use case?
It is mostly for small remote sites. The WAN link isn't good enough for them to come to the enterprise site at this time. So, we do a lot of file shares, VMs, etc. It's to run the local business.
Our FlexPods are NetApp FASs, Cisco UCS, and Cisco switches. That's our version of a FlexPod. We call them ROBOs (remote office/branch office). We have about a hundred throughout the world that we deploy in different regions. For us personally, I do the NetApp side of it. We're running NetApp version 9.5P6. That is the lowest version that we run in our ROBO environment.
While the deployment model is on-prem, we are moving to a backup model in the cloud for them for DR. In the next month or two, we are going to start that.
How has it helped my organization?
It's done really good things. A lot of it for us is being able to have that storage with the whole solution onsite at a small site, which may not have the WAN capabilities to use the corporate servers for their applications. So, that does help.
A lot of what we've done with the FlexPod is to replace hardware that was failing. We had a lot of UCS solutions go into replace IBM Blade Servers which were majorly failing. We had all types of problems with those.
We've also had challenges in the beginning where we didn't size sites right. We just totally blew it. We took their monthly closing down to a crawl, then ended up replacing it with an AFF solution, which was great. It really helped us out a lot.
It's just been a little bit here and a little bit there. The biggest thing is being able to have that remote site, and that they can keep running. If they lose the WAN, they can keep running. It's helped not having P1s and P2s at sites because they're dependent on corporate to be able to get something and they lose network connectivity. E.g., we had a site where the roof went. The site is in Fargo, North Dakota. They had a roof collapse at their site, but they kept going because, while they had other problems, they weren't reliant on going to a corporate data center to run their apps in the factory. They were sitting there able to keep continuously running even though they had a roof collapse.
We have done the all-flash at some sites. The one site where we totally blew the configuration, we came in with an All Flash FAS, and it went from them not knowing if they were going to be able to do year-end closing to year-end closing happening because they're an Oracle site. They had been on SAN previously, and all our ROBOs are NAS. We don't have any SAN in our ROBO environment, which is our FlexPod environment. So, they went from a SAN environment to a small FAS that didn't meet their needs, then with that AFF, we've had no problems since then. We installed it right before Christmas, literally two days before Christmas by pulling out the old and putting in the new.
For the entire stack, we have what we call a ROBO team in each of the regions. I'm part of the U.S. team. We have the same team work on this stack for every installation in the Americas, which includes places like North America, Mexico, and Brazil. It's really helped us because we've done documentation that we can push off to our separate teams that do the support, like server support, UCS support, and our storage support. This helps us out. Everything is the same. We've tried to keep everything the same and keep them as common as we can, so it helps with our operations team, which actually is in India. They know that if they can go to any one of those sites and there should be very similar setup.
For the longest time, with all the failures that they had with the IBM Blade Servers, our server staff was rushing to bring in storage and servers because of all the failures. Because of this solution, we now don't have very many problems. The only problems that we do have is sometimes storage gets a little out of control. They need more than they thought they needed. Other than that, it's been very smooth. We rarely have major problems at that size.
What is most valuable?
We've gotten it down to a science to install. So, it's been very easy to install. It has been very flexible for us because some sites don't need as much storage as other sites. Instead of going for a regular four terabyte, 12-drive solution, we can take it down to a two terabyte SaaS solution if the site doesn't need that much storage. Because we're trying not to have storage just sitting there, doing nothing, it's very flexible for us. We do have sites that have over a 100 terabytes. So, it's been a very flexible solution for us.
We do a little bit of Oracle at some of the sites, so the validated designs have been very good. We've had very good results. We have no complaints about latency or anything like that. Most of it is a lot of just file shares and stuff like that. But we do have Oracle and SQL at some sites.
What do I think about the stability of the solution?
What do I think about the scalability of the solution?
It is very scalable. It does depend on what model you get. For example, we don't try to put a small model in a site that we think would be growing.
How are customer service and technical support?
The technical support has been very good. I have had a few calls with them. I had one problem at the site where I had an aggregate that would not shrink after I had deleted some stuff. It took a few tries to get the right guy on the call. We do have a NetApp SAM with our company, and it really took getting to him to get the solution fixed.
Which solution did I use previously and why did I switch?
They were trying to replace all the older hardware with new hardware, getting some new sites as well. At some of the sites, they used the IBM Blade Servers, which were having high failure rates. That was a big wreck. We were going to a UCS solution, so they were trying to integrate into the UCS solution as well.
Three or four years ago, our management decided they were going to put in EMC VNX at a site that had a lot of Oracle in it. It was one of our bigger sites. They do big trucks there, and for the three years that VNX sat there, they had all types of Oracle problems in terms of latency issues, but could never get that latency issue fixed. We brought in a ROBO solution, and I didn't do any tweaking on it. I just put it in and put the Oracle on SaaS drives, then separated them out by themselves. We've had no complaints in two years.
How was the initial setup?
We did not use WWT for the initial setup, and we did have problems. A lot of it had to do with the gentleman who worked on the program left. From our perspective, it was a lot of trial and error. It took a couple deployments to get a rhythm to it. After that, since the first two to three deployments, it's been very smooth. With the same team, we know what we're doing. We have the same project leader.
What about the implementation team?
We did the deployment, but we did use our WWT. With WWT, we have them set up the basic configurations on everything. For the storage solution, they set up by the IPs and made sure everything is connected correctly. They don't get into the deep dive into the software or deployment. That is something we do.
They get it so when it's at the site, it gets plugged in. The network guy gets the ports plugged in and gets support set up. Then, we can get onto the storage and UCS, provision VMs, etc. Once that's setup, we can start working.
What was our ROI?
We have seen ROI.
The solution has decreased the unplanned downtime incidents in our organization by 25 to 30 percent.
What's my experience with pricing, setup cost, and licensing?
Everything is purchased, so we do not do any leasing with this product.
What other advice do I have?
I would rate the solution a solid nine (out of 10). The solution has been good for us. Nothing is perfect. That is why I wouldn't give it a ten. However, everything that we have done with it has been spot on. We've had very little problems with it. We're able to integrate it really well.
I would recommend going for this solution.
Which deployment model are you using for this solution?