What is most valuable?
We actually install Scality on the Apollo servers and so we have a ring, a Scality ring, where we store our customers' documents. That allowed us to migrate away from traditional NAS with a cost effective solution whose architecture is both scalable for the future and able to handle the PB scale of document content that we deal with.
How has it helped my organization?
Just not having to manage traditional NAS has made a big difference. Not having to manage traditional volumes and aggregates and LUNs and things like that. Being able to be flexible when it comes to that, and Apollo has made that possible.
What needs improvement?
We're pretty happy with the Apollo line of servers. It would be interesting to see the new hyper-converged DL380s. It would be cool to see if that type of same thinking about hyper-convergence was applied to the Apollo line of servers as well. It would be interesting, not on the storage-dense model of Apollo servers but on the compute-dense models of Apollo servers, to see kind of a hyper-converged solution running in those chassis that can have multiple compute nodes all in one. So that would be interesting to see if HPE could do something like that. It would make a compelling argument for them in their hyper-converged space. It would really complement the DL380 hyper-converged solution that they're providing now and would be I think a good choice for lots of people who are looking at hyper-converged.
What do I think about the stability of the solution?
We deployed our first ring on Apollo servers towards the end of last year so it's been running for eight or 10 months or so and it has had zero downtime.
What do I think about the scalability of the solution?
With the Apollo systems, we initially expected it to be of PB size. The great thing about the Apollo servers using Scality is that if we need to add more disks to those existing systems, that disk will instantly be usable to the ring. If we need to add more servers to have more compute power and more storage, we can do that as well.
How is customer service and technical support?
We've only contacted them to help replace drives when drives go bad as they do, but nothing beyond that.
Which solutions did we use previously?
So for a long time, we were storing our documents on a traditional NAS, through NetApp, and that got to the point where NetApp couldn't handle PB scale affordably. We're talking about tens of millions of dollars in order to buy a NetApp that could do PB scale on the number of IOPS that we needed. And on top of that, it was cost prohibitive to be able to scale out on traditional NAS, so the Apollo line became the clear choice, I guess. And deciding that we had to go to something like an object storage, that decision was made long before we decided on Apollo. It turns out that Apollo fit our decision to go to object store.
How was the initial setup?
The initial setup was pretty straightforward. The Apollo series that we use is basically the guts of a ProLiant DL380, which we've used many, many times in the past, but then allows us to put double the disk capacity of a traditional DL380 in that line of Apollo servers. And so setting it up was pretty easy because we've done Apollo servers in the past. The iLO functionality made it pretty straightforward and had no problems getting things deployed.
Which other solutions did I evaluate?
We spent a long time looking, actually, at doing the Scality ring on just commodity hardware from someone like Supermicro, and we found out that, in terms of reliability, supportability, ease of management, that having all our servers under the same contract through HPE, made the decision to use Apollo was apparently clear. Even though it was marginally a little bit more expensive up front, the total cost of ownership of having to manage those many servers was lower. This made the decision really easy.
What other advice do I have?
If someone came with a similar storage need, the Apollo servers do make a lot of sense, especially when you're talking about scale out object storage-type implementations. That Apollo line, it makes perfect sense from my perspective and I would recommend that.
Our first batch of Apollo servers that we got were so new that it was just hard to know kind of what to expect from HPE and what they wanted to deliver to us. The first batch of servers that we got were missing an iLO and that may have been a confusion between what we ordered and we thought we ordered or what we thought we would've had. But anyways, that way it was resolved quickly and the iLo modules were shipped out and there was no problem there. But just because it was so new when we first got it that there was just some speed bumps when we first ordered them. Otherwise, they're a very solid server.
Disclosure: I am a real user, and this review is based on my own experience and opinions.
Dec 09 2016