Tag Archives: HCI

3-Node Minimum? Not So Fast

For a long time, when you purchased HC3, you were told there was a 3 node minimum. This minimum of 3 nodes is what is required to create a resilient, highly available cluster. HC3 architecture, based on this 3 node cluster design, prevents data loss even in the event of a whole node failure. Despite these compelling reasons to require 3 nodes, Scale Computing last week announced a new single node appliance configuration.  Why now?

Recent product updates have enhanced the replication and disaster recovery capabilities of HC3 to make a single node appliance a compelling solution in several scenarios. One such scenario is the distributed enterprise. Organizations with multiple remote or branch offices may not have the infrastructure requirements to warrant a 3 node cluster. Instead, they can benefit from a single node appliance as a right-sized solution for the infrastructure.

Screen Shot 2016-07-18 at 2.06.52 PM

In a remote or branch office, a single node can run a number of workloads and easily be managed remotely from a central office. In spite of the lack of clustered, local high availability, single nodes can easily be replicated for DR back to an HC3 cluster at the central office, giving them a high level of protection. Deploying single nodes in this way offers an infrastructure solution for distributed enterprise that is both simple and affordable.

Another compelling scenario where the single node makes perfect sense is as a DR target for an HC3 cluster. Built-in replication can be configured quickly and without extra software to a single HC3 node located locally or remotely. While you will likely want the local high available and data protection a 3-node cluster provides for primary production, a single node may suffice for a DR strategy where you only need to failover your most critical VMs to continue operations temporarily. This use of a single node appliance is both cost effective and provides a high level of protection for your business.


Finally, although a single node has no clustered high availability, for very small environments the single node appliance can be deployed with a second appliance as a DR target to give an acceptable level of data loss and availability for many small businesses. The ease of deployment, ease of management, and DR capabilities of a full blown HC3 cluster are the same reasons to love the single node appliance for HC3.

Find out more about the single node appliance configuration (or as I like to call it, the SNAC-size HC3) in our press release and solution brief.

Screenshot 2016-07-13 09.34.07

What do DDOS Attacks Mean for Cloud Users?

Last Friday, a DDOS attack disrupted major parts of the internet in both North America and Europe. The attacks seems largely targeted on DNS provider Dyn disrupting access to major service providers such as Level 3, Zendesk, Okta, Github, Paypal, and more, according to sources like Gizmodo. This kind of botnet-driven DDOS attack is a harbinger of future attacks that can be carried out over an increasingly connected device world based on the Internet of Things (IoT) and poorly secured devices.


This disruption highlights a particular vulnerability to businesses that have chosen to rely on cloud-based services like IaaS, SaaS, or PaaS. The ability to connect to these services is critical to business operations and even though the service may be running, if users cannot connect, it is considered downtime.  What is particularly scary about these attacks for small and midmarket organizations especially, is that they become victims of circumstance from attacks directed at larger targets.

As the IoT becomes more of a reality, with more and more devices of questionable security joining the internet, the potential for these attacks and their severity can increase. I recently wrote about how to compare cloud computing and on-prem hypercoverged infrastructure (HCI) solutions, and one of the decision points was reliance on the internet. So it is not only a matter of ensuring a stable internet provider, but also the stability of the internet in general with the possibility of attacks targeting a number of different services.

Organizations running services on-prem were not affected by this attack because it did not affect any internal network environments. Choosing to run infrastructure and services internally definitely mitigates the risk of outage from external forces like collateral damage from attacks on service providers. Many organizations that choose cloud services do so for simplicity and convenience because traditional IT infrastructure, even with virtualization, is complex and can be difficult to implement, particularly for small and midsize organizations. It has only been recently that hyperconverged infrastructure has made on-prem infrastructure as simple to use as the cloud.

The future is still uncertain on how organizations will ultimately balance their IT infrastructure between on-prem and cloud in what is loosely called hybrid cloud. Likely it will simply continue to evolve continuously with more emerging technology. At the moment, however, organizations have the choice of easy-to-use hyperconverged infrastructure for increased security and stability, or choose to go with cloud providers for complete hands-off management and third party reliance.

As I mentioned in my cloud vs. HCI article, there are valid reasons to go with either and the solution may likely be a combination of the two. Organizations should be aware that on-prem IT infrastructure no longer needs to be the complicated mess of server vendors, storage vendors, hypervisor vendors, and DR solution vendors. Hyperconverged infrastructure is a viable option for organizations of any size to keep services on-prem, stable, and secure against collateral DDOS damage.

Screenshot 2016-07-13 09.34.07

IT Infrastructure: Deploy. Integrate. Repeat.

Have you ever wondered if you are stuck in an IT infrastructure loop, continuously deploying the same types of components and integrating them into an overall infrastructure architecture? Servers for CPU and RAM, storage appliances, hypervisor software, and disaster recovery software/appliances are just some of the different components that you’ve put together from different vendors to create your IT infrastructure.mobius

This model of infrastructure design, combining components from different vendors, has been around for at least a couple decades. Virtualization has reduced the hardware footprint but it added one more component, the hypervisor, to the overall mix. As these component technologies like compute and storage have evolved within the rise of virtualization, they have been modified to function but have not necessarily been optimized for efficiency.

Take storage for example.  SANs were an obvious fit for virtualization early on. However, layers of inefficient storage protocols and virtual storage appliances that combined the SAN with virtualization were not efficient. If not for SSD storage, the performance of these systems would be unacceptable at best. But IT continues to implement these architectures because it has been done this way for so long, regardless of the inherent inefficiencies. Luckily, the next generation of infrastructure has arrived in the form of hyperconvergence to break this routine.

Hyperconverged infrastructure (HCI) combines compute, storage, virtualization, and even disaster recovery into a single appliance that can be clustered for high availability.  No more purchasing all of the components separately from different vendors, no more making sure all of the components are compatible, and no more dealing with support and maintenance from multiple vendors on different schedules.

Not all HCI systems are equal, though, as some still adhere to some separate components.  Some use third party hypervisors that require separate licensing. Some still adhere to SAN architectures that require virtual storage appliances (VSAs) or other inefficient storage architectures requiring excessive resource overhead and require SSD storage for caching to overcome inefficiencies.

Not only does HCI reduce vendor management and complexity but when done correctly,  embeds storage  in the hypervisor and offers it as a direct attached, block access storage system to VM workloads. This significantly improves the I/O performance of storage for virtualization. This architecture provides excellent performance for spinning disk so when SSD is added as a second storage tier, the storage performance is greatly improved.  Also, because the storage is including in the appliance, it eliminates managing a separate SAN appliance.

HCI goes even further in simplifying IT infrastructure to allow management of the whole system from a single interface.  Because the architecture is managed as a single unit and prevalidated, there is no effort spent making sure the various components work together. When the system is truly hyperconverged, including the hypervisor, there is greater control in automation so that software and firmware updates can be done without disruption to running VMs.  And for scaling out, new appliances can be added to a cluster without disruption as well.

The result of these simplifications and improvements of infrastructure in hyperconvergence is an infrastructure that can be deployed quickly, scaled easily, and that requires little management. It embodies many of the benefits of the cloud where the infrastructure is virtually transparent.  Instead of spending time managing infrastructure, administrators can focus time managing apps and processes rather than hardware and infrastructure.

Infrastructure should no longer require certified storage experts, virtualization experts, or any kind of hardware experts. Administrators should no longer need entire weekends or month-long projects to deploy and integrate infrastructure or spend sleepless nights dealing with failures. Hyperconvergence breaks the cycle of infrastructure as a variety of different vendors and components. Instead, it makes infrastructure a simple, available, and trusted commodity.

Screenshot 2016-07-13 09.34.07

Scale Computing Keeps Storage Simple and Efficient

Hyperconvergence is the combination of storage, compute, and virtualization. In a traditional virtualization architecture, combining these three components from different vendors can be complex and unwieldy without the right number of experts and administrators. When hyperconverged into a single solution, the complexity can be eliminated, if done correctly.

At Scale Computing we looked at the traditional architecture to identify the complexity we wanted to eliminate. The storage architecture that used SAN or NAS storage for virtualization turned out to be very complex.  In order to translate storage from the SAN or NAS to a virtual machine, we counted 7 layers of object files, file systems, and protocols through which I/O had to traverse to go from the VM to the hardware. Why was this the case?

Because the storage system and the hypervisor were from different vendors, and not designed specifically to work with each other, they needed these layers of protocol translation to integrate. The solution at Scale Computing for our HC3 was to own the hypervisor (HyperCore OS) and the storage system (SCRIBE) so we could eliminate these extra layers and make storage work with VMs just like direct attached storage works with a traditional server. I call it a Block Access, Direct Attached Storage System because I like the acronym.

Why didn’t other “hyperconverged” vendors do the same? Primarily because they are not really hyperconverged and they don’t own the hypervisor. As with traditional virtualization architectures, the problem of the storage and hypervisor being from different vendors prevents efficiently integrated storage for VMs.  These are storage systems being designed to support one or more third party hypervisors. These solutions generally use virtual storage appliances (VSAs) with more or less the same storage architecture as the traditional virtualization I mentioned earlier.

VSAs not only add to the inefficiency but they consume CPU and RAM resources that could otherwise be used by VM workloads.  To overcome these inefficiencies, these solutions use flash storage for caching to avoid performance issues. In some cases, these solutions have added extra processing cards to their hardware nodes to offload processing. Without being able to provide efficient storage on commodity hardware, they just can’t compete with the low price AND storage efficiency of the HC3.

The efficiency of design for HC3 performance and low price is only part of the story. We also designed the storage to combine all of the disks in a cluster into a single pool that is wide striped across the cluster for redundancy and high availability.  This pooling also allows for complete flexibility of storage usage across all nodes.  The storage pool can contain both SSD and HDD tiers and both tiers are wide striped, highly available, and accessible across the entire virtualization cluster, even on nodes that may have no physical SSD drives.

To keep the tiering both simple and efficient, we designed our own automated tiering mechanism to automatically utilize the SSD storage tier for the blocks of data with the highest I/O.  By default, the storage will optimize the SSD tier for the best overall storage efficiency without anything to manage. We wanted to eliminate the idea that someone would need a degree or certification in storage to use virtualization.

We did recognize that users might occasionally need some control over storage performance so we implemented a simple tuning mechanism that can be used to give each disk in a cluster a relative level of SSD utilization priority within the cluster. This means you can tune a disk up or down, on the fly, if you know that disk requires less or more I/O and SSD than other disks. You don’t need to know how much SSD it needs, but only that it needs less or more than other disks in the cluster and the automation takes care of the rest.  We included a total of 12 levels of prioritization from 0-11 or no SSD and scaling to 11 for putting all data on SSD if available.

Screenshot 2016-04-19 13.07.06

The result of all of the design considerations for HC3 at Scale Computing is simplicity for efficiency, ease of use, and low cost. We’re different and we want to be. It’s as simple as that.

Screenshot 2016-07-13 09.34.07