What is our primary use case?
We are using a huge portion for NAS as the main use case for Qumulo, and we are also evaluating the S3 option.
We have a lot of cloud-native workload on Qumulo, which impacts our global collaboration efforts. We have ML workload, AI workload, and applications hosted on the cloud. We have some customer-facing applications as well, and we also utilize snapshot ability for some of the applications that actually use the ongoing on-cloud snapshot availability. We have quite a lot of cloud-native applications hosted.
How has it helped my organization?
The benefit of Qumulo's multi-protocol compatibility for integrating into existing workflows is pretty good. We moved a lot of our data from NetApp to Qumulo. Before we had Qumulo, we were using Cohesity NAS, so a lot of movement was actually done from Cohesity NAS, which is also one of the software-defined NAS that we had, and we moved a lot of workload from NetApp NAS as well. The migration was easy; we had a huge volume of data, but we were able to complete it with little to no problems.
The key benefits that I have seen from using Qumulo are performance and scalability. We get pretty decent performance and do not actually have to deal with the overhead of an appliance. We are just using hardware from Dell with Qumulo's guest OS.
Qumulo is software-defined, so it has a pretty light UI, and we do actually have a lot of expansion going on. We have snapshot ability in Qumulo as well, which is a pretty awesome feature to have. We have been evaluating the anti-ransomware ability of Qumulo as well, and I think we have done some testing in the QA environment that went pretty well. We also have some protection against malware. There isn't a lot of trade-off; in software, in technology, or anything, there is always a trade-off, but I think that's negligible given the kind of performance that we get.
What is most valuable?
What I appreciate about Qumulo is that it's software-defined, easy to deploy, easy to procure, easy to expand, and easy to scale, along with quite a lot of other things. It's not an appliance, so upgrades are easier, and I think so far we have had good feedback from the consumers who have been using it.
We have a lot of AI workload which is hosted on Qumulo, and there are no challenges. Of course, there are small issues here and there but nothing big. If we consider the performance also, we get pretty decent performance compared to the other software-defined NAS that we have available on the premises.
All in all, it's a very good product, resilient, with good performance and scalability that is pretty easy.
What needs improvement?
The key differences, both pros and cons of Qumulo in comparison to other technologies are that I have outlined quite a lot of pros, while the cons I can see include some challenges between the software and hardware. We have hardware that is from Dell, and software from Qumulo, so sometimes there is a synchronization issue. If we get into an issue, hardware says, 'Talk to software, Qumulo,' and Qumulo says, 'Talk to Dell.'
There is this bit of a mismatch, and I think Dell and Qumulo, or probably some other hardware vendors like Intel or HP, should have more collaborative relations so users do not end up having a lot of issues. Another thing is that I think the UI sometimes has issues during the upgrade. We immediately open an incident, and then engineers are assigned. These small issues can be considered as cons.
There are no missing features in Qumulo that I would like to see included.
For how long have I used the solution?
I have been working with Qumulo for almost a year. It's a pretty new product in the company that I'm working with, so it's a fairly new product, but approximately one year.
What do I think about the stability of the solution?
We do get performance issues from time to time with Qumulo, but we get support. A lot of performance issues also depend on how the workload is handled. Suppose we put additional workload on the product, and the product cannot cater to it, of course, we will get into performance issues. For instance, if a client is making 100 million API calls, any product can encounter issues.
We are also working on some QoS and setting some quotas so we can limit how the end-user applications behave, and we're coming up with ideas to control mechanisms on the end-users.
What do I think about the scalability of the solution?
We are leveraging Qumulo's scalable file system.
As of now, we have a footprint of almost 100 petabytes, and the team is actually in discussion with Qumulo team as we are in the process of procuring more. By the end of 2027, we are expecting at least a 50% organic data growth. Of course, that includes some metadata and some replication data as well, but we could very well hit 150 petabytes in the coming year, by 2027 end.
How are customer service and support?
I would evaluate Qumulo's technical support team as pretty good. We have a dedicated engineer who is available 24/7 whenever we raise an issue. It's very good support, and we have a very good relationship with Qumulo engineers.
The support team is always available, even during the weekend. We've had quite a few issues where we needed support over the weekend, and I would say within the defined SLA, we were allocated an engineer from Qumulo.
I would rate Qumulo's tech support as a 9 out of 10.
Which solution did I use previously and why did I switch?
The reason I decided to switch to Qumulo is that it was comparatively a lot less money than NetApp.
How was the initial setup?
The initial setup worked well because we got support from Qumulo, and as far as the hardware procurement, that is a different story. Racking and cabling were done by our internal engineers, and the cluster build did not take too much time.
The licensing had a bit of a challenge, but we got support during migration when we moved the workflow from another product to Qumulo. That was when we faced major issues, and Qumulo support was actually pretty helpful in that scenario.
Which other solutions did I evaluate?
I have evaluated other solutions available in the market, including Pure and some of the Dell software-defined options, and I also evaluated Lightbits. Ultimately, we decided to go with Qumulo.
What other advice do I have?
If you have AI workload, I think Qumulo is a pretty good bet. If you have ML workload, it's pretty good as well. However, if you are using a database, that isn't the product to choose. Structured data is not appropriate for Qumulo. For file-related work, absolutely proceed with it. Anything related to rows and columns is not suitable for this product, so don't go in that direction.
Which deployment model are you using for this solution?
On-premises