Scale Computing HC3 - Excellent and Unique Hyper-Converged Platform with a few Rough Edges
Updated December 23, 2017
Scale Computing HC3 - Excellent and Unique Hyper-Converged Platform with a few Rough Edges
Score 7 out of 10
Vetted Review
Verified User
Overall Satisfaction with Scale Computing HC3
Scale HC3 is being used to virtualize a large portion of our physical IT stack.
We purchased it for the following reasons:
1. Simplify administration of our infrastructure.
2. Provide out-of-the box High-Availability and/or near High-Availability for critical systems
3. Decrease TCO for our physical infrastructure
4. Simplify our infrastructure by collapsing all of the requirements for a clustered virtualization solution down into one software/hardware stack, including compute, storage, and networking.
5. Decrease deployment time for new environments. Using near-instant thin-cloning we can deploy multiple systems in a very short amount of time.
6. SCALE provided a solid platform alternative to other traditional high-availability solutions, like SQL Clustering. In this case we simplify life by deploying SQL server's as stand-alone but because they are virtual and on a Hyper-Converged platform, they are HA by default. So we get a lot of the benefits of SQL clustering without the headache and complexity of actual SQL clustering.
We purchased it for the following reasons:
1. Simplify administration of our infrastructure.
2. Provide out-of-the box High-Availability and/or near High-Availability for critical systems
3. Decrease TCO for our physical infrastructure
4. Simplify our infrastructure by collapsing all of the requirements for a clustered virtualization solution down into one software/hardware stack, including compute, storage, and networking.
5. Decrease deployment time for new environments. Using near-instant thin-cloning we can deploy multiple systems in a very short amount of time.
6. SCALE provided a solid platform alternative to other traditional high-availability solutions, like SQL Clustering. In this case we simplify life by deploying SQL server's as stand-alone but because they are virtual and on a Hyper-Converged platform, they are HA by default. So we get a lot of the benefits of SQL clustering without the headache and complexity of actual SQL clustering.
Pros
- Scale has focused on ease-of-management and it shows. Their UI is incredible and easily the best in the industry.
- SCRIBE - the underlying storage layer is brilliant and if you are a techie, worth taking a closer look. The way SCALE has done storage for hyper-converged is very different from almost all of the other major players in a very good way. SCRIBE VM snapshots and cloning is incredible, very fast, very efficient.
- Simplified licensing model - NO VMWARE - this is a big deal. Scale has built-the-entire-stack from almost the ground up. You only have one company to deal with when it comes to licensing your hyper-converged stack. Furthermore, all SCALE HC3 clusters have ALL FEATURES. There is no nickel-and-dime (and arm-n-leg) for additional functionality.
Cons
- I think the company has been focused on the SMB market and as a result are not familiar with dealing with the issues and concerns of either larger companies/deployments and/or companies like ours which service larger clients. We have no tolerance for unscheduled downtime. When it comes to dealing with support they are mostly all very knowledgeable but they seem a bit green when it comes to understanding how "serious" uptime is for their customers that are running mission-critical systems.
- I think the company has been focused on the SMB market, and as a result I get the impression they are not overly familiar with companies that run mixed and/or more highly-demanding workloads. We are pushing our cluster hard. We run a heavy mix of both virtualized SQL and WEB tier systems. Based on multiple interactions with the company, I don't think they have very many customers operating in this space and as a result, the platform hasn't been as well vetted based on these kinds of workloads. That said, I realize that virtualizing multiple busy database servers on a single platform is huge ask. All of the above being said... we have not had any system failures that have impacted any of our production VM's and by-and-large the platform is pretty darn solid. We are seeing some performance hits, particularly around storage, as our cluster gets more busy however we haven't done wide-spread performance testing to get an accurate measurement and I believe scribe is dynamically tiering data to keep our busiest systems running effectively. I guess I would like to just see a higher "low watermark" for storage performance on VM's.
- Support is overall very responsive however in-depth issues can drag-on a bit, which is understandable.
- They are still missing some features such as role-based administrative access which are important to, once again, larger companies and/or those who service larger companies.
- All-in-all - Scale has historically targeted SMB - and as a result they have a ton of core strengths that are unique to their company and their product. However, in focusing on this market segment, I think they have a bit of tunnel-vision which has resulted in them missing some crucial areas for businesses that take their IT very serious because it is a critical part of their service delivery/day-to-day operations.
- We have virtualized ~25 systems at this point and dropped two older SAN arrays from our environment. This has drastically reduced administrative complexity in our environment.
- Things are just "more stable" - we were often putting out fires with failing hardware across multiple systems. With scale, we have one platform which is highly-redundant and this allows me to finally get some sleep at night. Sleep is priceless.
- For new projects, we have cut new environment deployment time drastically. Setting up a new hosted environment for a project used to take 2 - 4 weeks, it now takes 2 - 4 days. It also cuts down expense. When we need dedicated environments we now spin-up VM's on a platform that is already budgeted and paid for vs. having to buy new dedicated hardware.
- An Active-Passive SQL cluster with SAN storage can run anywhere from $20K (assuming you already have the SAN) --> $60k or even $100k if you need to put in the SAN and supporting infrastructure. A SCALE cluster however can support multiple SQL servers and provide almost the same level of redundancy/high-availability with drastically lower cost and complexity vs a traditional cluster.
I don't have any real-world experience with Nutanix but we did demo their product and I did a fair bit of research. I also looked heavily at other products in the hyper-converged space.
My "end of the day" analysis is that SCALE HC3 is one of the few "full stack" platforms. What I mean by that is that most other products are just some re-mix of existing storage technologies (SAN) that have been virtualized and put into the hyper-visor stack. As a result, the management of the system is fragmented. HC3 on the other hand was built "from the ground up" to be a Hyper-Converged product and as a result, everything is managed under one roof. You don't have to "worry" about 10 different components all being configured correctly to work together, SCALE is a single, unified product and that makes them different than 90% of the other players out there which are pretty much borrowing VMware and then tacking on some kind of VSAN to make a Hyper-Converged stack. This might sound intangible but it results in real-world differences in how the product works and is managed on a day-to-day basis.
My "end of the day" analysis is that SCALE HC3 is one of the few "full stack" platforms. What I mean by that is that most other products are just some re-mix of existing storage technologies (SAN) that have been virtualized and put into the hyper-visor stack. As a result, the management of the system is fragmented. HC3 on the other hand was built "from the ground up" to be a Hyper-Converged product and as a result, everything is managed under one roof. You don't have to "worry" about 10 different components all being configured correctly to work together, SCALE is a single, unified product and that makes them different than 90% of the other players out there which are pretty much borrowing VMware and then tacking on some kind of VSAN to make a Hyper-Converged stack. This might sound intangible but it results in real-world differences in how the product works and is managed on a day-to-day basis.
Comments
Please log in to join the conversation