What is our primary use case?
We use FlexPod in our data centers. We serve all of our infrastructure off of it, which includes Exchange, SQL, SharePoint, and Citrix. It is all virtualized. We are also using the file share from FlexPod with Snapshotting and SnapMirroring for disaster recovery (DR) between data centers.
How has it helped my organization?
We haven't experienced any data loss while on NetApp. The stability of it has probably been the biggest benefit. Because of FlexPod's performance and flexibility, our company is doing much better than what we previously used.
We found FlexPod to be innovative when it comes to compute, storage, and networking. We've taken advantage of their storage optimizations to obtain better use out of the space. We upgraded to All Flash FAS (AFF), which has provided a huge performance increase that we haven't barely scratched the surface of. We have plenty of overhead, so that's always nice when taking on tasks which might have otherwise taxed a smaller system. However, we have a lot of overhead, so this isn't an issue for us.
Because of the stability that we have had on it, it has met our needs on everything. We haven't had a shortcoming in performance nor data loss.
What is most valuable?
In regards to DR and backup:
What needs improvement?
Validate designs are hard. They don't validate all of the available options. We don't generally end up in a validated configuration. We did on our initial install when they first rolled out the FlexPod platform. Over time, we've done upgrades, and we don't necessarily fit into a validated design anymore.
We would like them to improve the validate designs. It is hard to stay in a supported config with the software and firmware versions of the platform. It's always a concern to ensure things not only work well, but they work at all. If we run into incompatibility inside of the NetApp, Cisco, or VMware versions, it can cause real issues.
They should continue to educate and support their Tier 1 support, so we have better, faster resolutions. As the years have gone by, we haven't quite received as good resolution at Tier 1 as we used to. Occasionally, scheduling techs onsite is problematic. There are some gaps in the handoff between the call-in support to on-site support. It would be nice if this was cleaned up, so we didn't have to be quite as involved with verifying techs will be on site or ensuring that techs onsite receive all the information.
What do I think about the stability of the solution?
As long as it stays in a supported config, the stability is very good. If you leave the supported config, you get directed to come back into a supported config if you have any issues.
We have good resiliency with our FlexPods. I don't know if we've taken advantage of the built-in HA.
What do I think about the scalability of the solution?
We haven't had much experience with scalability. We gave ourselves room to grow into the product. We've only done any real scaling at refresh time.
How are customer service and technical support?
It has not always been the single point of contact for all of vendors who participate as it was sold to be. Occasionally, we end up having to go to each vendor, and there isn't as much cross-vendor support as we had wanted.
There is always room for improvement in support. We want the intercompany communications to not have us have to contact vendors separately to work on one issue. We want them to own it internally, which would be a lot more helpful. This is what they're supposed to do.
Compared to some other vendors, we still receive good support. Unfortunately, the issue being that they still seem to be separate support buckets rather than integrated support. It's hard to ding the platform overall, but that's probably where I would ding it at the moment.
Which solution did I use previously and why did I switch?
We were using HPE G-Series Servers. We needed a lot more space and performance, since I'm not sure that we had good performance metrics at the time that we moved solutions. However, we were looking to expand our Exchange environment and have more SQL. We wanted making sure that we had enough I/O, and the FlexPod system had it. In addition, integrating with UCS made it much more flexible to add compute in our VM environment, and we were going from physical to virtual at the time. Thus, we cut down on the amount of space and power that we were using by going to blade chassis.
How was the initial setup?
The initial setup was complex at the time our deployment where there was a lot of moving parts. My understanding is that they have since tried to implement more streamlining.
What about the implementation team?
We used Plan B Technologies out of Maryland, and we also used NetApp. We had a good experience with the install. It was all-new moving parts for us, since FlexPod was brand new at the time. We spent a fair amount of time whiteboarding the solution with them. We visited Raleigh-Durham to go on campus to see some of the hardware to get a better understanding of what we were going to be buying.
What was our ROI?
We have seen an improvement in application performance. We are pushing a lot more I/O and flexibility. We came from systems which did not have thin provisioning. Therefore, we are more flexible in being able to give out space or have I/O, especially with the AFF being all-flash.
Which other solutions did I evaluate?
We probably looked at Dell EMC. We were on HPE SAN for quite awhile. I don't know if we looked at anybody else.
One of the deciders for us in looking at NetApp was, even years ago, they just seemed to be in a much better position in the marketplace. We were pretty confident that they would be around in five years, whereas, some of the other smaller vendors might not be, especially with consolidations going on.
What other advice do I have?
We have saved time with Snapshots, SnapMirrors, and backup and DR capabilities versus other platforms that we have looked at in the past. However, for new deployments, we have not saved, because we don't have any automation on top for deploying VMs or shares. It doesn't really seem to be part of the FlexPod platform.
We don't use it for hybrid cloud, multi-cloud environments, or Managed Private Cloud.
Everything that we are looking for feature-wise seems to be coming out in ONTAP or VMware releases.