HCI

by

IT professionals often reject hyperconverged infrastructure (HCI) because they want to integrate HCI into their traditional three-tier architecture, and most HCI solutions can’t meet this requirement. HCI vendors also only focus on three tiers of the typical data center: networking, virtualization, and storage. However, a fourth tier is far too often overlooked: data availability and protection.

VergeOS’ Ultraconverged Infrastructure (UCI) can integrate into a customer’s traditional three-tier architecture. In this article, we will cover how VergeOS enables a more gradual onramp to UCI instead of a complete overhaul on day one and allows you to use components from each of your existing tiers.

UCI, an HCI Solution You Actually Want

Integrate HCI into a Three-Tier Architecture

The concept of HCI has merit. Take the data center tiers running on dedicated, vendor-provided hardware and reenable them as software, converging that software onto a single commodity server, liberating the organization from vendor lock-in and markups. The problem faced when IT attempts to integrate HCI into a traditional three-tier architecture is that HCI doesn’t truly converge anything. HCI stacks these tiers as software packages on top of each other, essentially recreating the same technology stack within a single server. It also forces them to exclude external SANs and existing servers.

UCI changes this by collapsing these stacks into a single, cohesive operating environment that is significantly more efficient. This elegant code base reduces overhead by as much as 30% while improving performance. The result is a VMware Alternative that is more portable, performs better, and provides improved data resilience.

UCI, an HCI Solution That Supports Fibre Channel SANs

One of the largest investments in the data center is the storage tier. If you have hundreds of thousands of dollars invested in a fibre-channel (FC) storage area network (SAN), you will want to integrate HCI into your traditional three-tier architecture so that you can continue to benefit from your significant investment. The problem is that most HCI solutions don’t support them. Because VergeIO owns all the code within VergeOS at a very deep level, we have taken the necessary steps to support fibre-channel-attached storage.

Click to Register

When using VergeIO and a FC SAN, you’ll need a physically local drive in each server to load VergeOS and store metadata, but the VM data can reside on the FC array. You’ll create a LUN as a virtual drive for each node that contributes to storage for the VergeOS instance. The VergeOS will manage and aggregate virtual drives into a pool that the VMs can use for data storage. VergeOS will also provide an additional layer of protection from hardware failure and add a deduplication capability. If your array already has deduplication, you can turn it off or leave it on; we’ve been unable to measure any noticeable performance impact.

In the future, when it is time to add additional capacity to the environment, you have options. First, you can continue to add capacity to your existing FC array, or you can take advantage of the 10X cost savings of using internal server class NVMe SSDs and/or HDDs. Also, as new storage technologies come to market, there is a much higher likelihood of them being supported sooner under VergeOS than waiting for the storage vendor to retrofit their designs to support them.

UCI, an HCI Solution That Supports Blade Servers

VergeOS is unique in that it can run a variety of different servers within the same instance. You can mix different server manufacturers, processor generations, and even processor manufacturers (Intel and AMD). You can also mix different server configurations. Some servers can be storage-heavy, and others can be processor-heavy.

Integrate HCI into a Three-Tier Architecture

Blade servers are processor-heavy nodes because they can only support a limited amount of internal storage. This makes blade servers problematic for IT professionals looking to integrate HCI into a traditional three-tier architecture. It forces them to reconsider the server strategy, which blade server customers are reluctant to do.

To incorporate a blade server into the VergeIO instance, you need to have either one internal drive store to boot VergeOS, or VergeOS also supports PXE booting, so an internal drive is not required. From there, the blade server can access its data from the shared pool of storage that VergeOS creates from the other nodes, or in the aforementioned FC SAN. In fact, if the blade server is FC equipped, it can “contribute” storage to the overall instance by assigning it a LUN as described above. Hosting storage does require slightly more RAM, so it is best to review the exact configuration with our technical architects.

As is the case with storage, VergeOS brings added flexibility for server expansion or refreshes. When it comes time to upgrade or expand your server infrastructure, you can continue to add blade servers or use standard 1U, 2U, or larger commodity servers of your choice.

UCI, an HCI Solution That Supports External Networks

Integrate HCI into a Three-Tier Architecture

VergeOS has a robust set of layer 2 and layer 3 networking functionality, but its use is optional. If you’ve invested in one of the market leaders for networking functionality, VergeOS can easily interoperate with them. As with the above, as time goes on, you can decide to use some of the network functionality built into VergeOS, saving you the cost of additional dedicated appliances, or you can continue to use your current vendor. To learn more about VergeOS, watch our Networking Fundamentals video.

UCI, an HCI Solution That Supports Third-Party Backup

VergeOS’ Data Availability Services combined with ioGuardian capabilities lessen IT’s dependency on the backup infrastructure for most recovery efforts, and many customers decide that is all the protection they need. VergeOS provides an in-guest agent that can quiesce when applying consistent snapshots. Still, some customers want to use third-party backup solutions to meet these requirements:

Copy of Last Resort

VergeOS can export the VMs’ raw files via an NFS mount point that almost any backup product can browse to meet compliance or vendor-independent data copy requirements to a different storage system, even tape.

File Level Restoration

As we show in this video, several types of file-level restoration can be done by leveraging VergeOS snapshots. Customers who want a centralized file-level recovery can install an in-guest agent from their backup software application into the virtual machines that are likely to require single-file recoveries.

Again, most customers find they can meet 100% of their data protection and recovery needs using VergeIO. Still, its support of third-party backup products enables IT to continue to use these solutions if required, or until they gain full confidence in VergOS’ capabilities.

Conclusion

Schedule a Technical Deep Dive on VergeOS

VergeOS’ Ultraconverged Infrastructure (UCI) effectively addresses the challenges of integrating HCI into a traditional three-tier architecture by supporting diverse hardware, including blade servers and FC SANs, and offering robust networking capabilities and data protection. As a versatile VMware alternative, VergeOS truly converges data center tiers into an efficient environment, allowing organizations to leverage existing investments and adapt to future technologies. This combination of compatibility, enhanced performance, and cost savings makes VergeOS’ UCI a very practical choice for modernizing IT infrastructure.

Filed Under: Storage Tagged With: , ,

by

Many IT Professionals disqualify Hyperconverged Infrastructure (HCI) as a VMware Alternative because they feel that overcoming HCI shortcomings is more expensive than continuing to use legacy three-tier architectures. These shortcomings are in areas where HCI was initially intended to excel: price, performance, scalability, and simplicity.

Overcoming HCI shortcomings led the VergeIO team to start from a clean slate and create the industry’s first UltraConverged Infrastructure (UCI) solution, VergeOS. It provides a superior, cost-effective alternative to HCI and the traditional three-tier infrastructure. Watch our on-demand webinar, “Beyond HCI” for a comparison of VergeOS’ UCI to the VMware and Nutanix HCI solutions.

In this article, we will explore the critical shortcomings of HCI, which lead IT planners to continue to leverage dedicated infrastructure, and how UCI overcomes them.

Can HCI Deliver Better CPU Performance?

The short answer is no. HCI can’t outperform the traditional three-tier architecture, which has dedicated hardware powering each tier. The reason is that HCI uses shared hardware to power at least four separate software packages, to deliver:


1) A package that contains the core operating system and the hypervisor.
2) A package that contains the storage software.
3) A package that contains the network software.
4) A package that contains the management software.

Each of these packages consumes CPU resources, and HCI’s lack of integration between them means it discards the potential gains in efficiency that combining them into a single code base would deliver. IT must install these packages as separate entities, making installation more complex. Then, once they are all installed and running and IT is ready to create a VM for the organization instead of for infrastructure, they’ve already lost as much as 20% of the CPU resources. To compensate, HCI requires the purchase of more powerful (and expensive) servers.

Another challenge is that as your application executes within the VM, it is probably running through much of the above stack. It uses the CPU to process requests from users. It is using storage IO to read and write data. It is using network resources to receive that input and deliver results. Finally, the VM’s health is reported to a management console in order to report on its health. Each transaction that the application executes is mired down in overhead.

The inefficiency of the code base forces HCI vendors into a predictable pattern of making customers buy turnkey hardware and software solutions from them, or buy new hardware using a strict hardware compatibility matrix. The required hardware must also be overpowered to compensate for the inefficiency, increasing costs.

UCI Delivers Near Bare Metal Performance

Can HCI Deliver Better CPU Performance

The fundamental difference between UCI and HCI is that UCI goes the extra step and eliminates the four separate software modules listed above. Instead, it integrates them into a cohesive code base, increasing resource efficiency. VergeIO typically uses less than 3-5% of CPU resources. The efficiency of resources also improves each VM transaction since no layers of code are involved in each request.

VergeIO customers consistently report measurable improvement in performance-demanding VMs and increased VM density while using existing hardware. Many VergeIO customers even report virtualizing formerly bare metal workloads and seeing a performance improvement.

Can HCI Deliver Better Storage Performance?

Concerns over storage performance are the number one reason customers will disqualify HCI in favor of dedicated three-tier architectures. Storage performance and storage scalability are very legitimate concerns for HCI vendors. First, most HCI vendors don’t use their own storage software. Often, they use an open-source solution like ZFS or CEPH, which were not designed for the unique requirements of providing storage services to a virtualized infrastructure. The solution remains convoluted even if they have their storage code.

Because of the lack of integration, HCI vendors have stumbled through implementing advanced drive failure protection and storage efficiency algorithms like deduplication. Adding these capabilities post-facto adds another layer to an already complex combination of software. For this reason, most vendors force customers to choose between storage efficiency and advanced drive redundancy.

Can HCI Deliver Better Storage Performance

UCI Delivers Better Storage Performance

The storage performance delivered by a UCI solution like VergeOS can outperform an HCI solution and a dedicated storage array while significantly reducing the storage cost. VergeOS correctly balances storage efficiency and storage performance. Its deep integration into the core software enables features like drive failure protection and global inline deduplication to work without adversely impacting performance.

Can HCI Deliver Better Scale?

By its very nature, HCI is scale-out, so scale should be an advantage, but once again, it falls short. Most HCI solutions require three servers (nodes) to start. It forces many small data centers to use two servers and a SAN or NAS. Many of these customers would benefit from a simple two-node solution that includes all the storage and networking functionality within those nodes.

HCI also doesn’t meet the scaling demands of enterprises. Most HCI solutions can only scale to eight nodes per instance before network traffic becomes challenging. They also have to buy similar nodes with each upgrade. If their needs change, they have to start an entirely separate instance of the HCI environment. As a result, many customers opt for the legacy three-tier architecture because each tier can be scaled independently of the other.

UCI Delivers Better Scale

Overcoming HCI shortcomings requires an infrastructure that is flexible and can adapt to the changing demands of the organization.

As we explain in our article “The Full Value of Scale”, a UCI solution like VergeOS delivers a three-dimensional scale. It can start with as few as two nodes, making it ideal for small data centers and remote offices, but it can also scale to hundreds of nodes to meet the needs of the most demanding enterprise. Nodes within the VergeOS instance can be different from each other. Customers can use nodes that provide balanced compute and storage, or mainly compute, or mostly storage, have GPUs installed, or any combination of the above.

Why is HCI More Expensive than Legacy Three Tier?

HCI should have a significant price advantage over legacy three-tier infrastructure. By definition, it is supposed to use off-the-shelf commodity servers, storage, and network hardware. However, this is seldom the case. Under the guise of “making it easy to install and support,” these vendors either require you to buy a turnkey hardware and software solution from them, require you to buy a specific configuration from one of their “certified” hardware vendors, or have a rigorous hardware compatibility list.

The inefficiency of layering virtualization software, storage software, and networking software as three separate software packages also drives up the cost of HCI. Customers must buy much more powerful nodes to support the additional overhead.

UCI Delivers Better TCO and ROI

Overcoming HCI shortcomings requires an infrastructure that can deliver the promise of reduced cost and simplified operations.

VergeOS’ tight integration of the hypervisor, storage, and networking software means customers can actually use off-the-shelf storage from whichever vendor they choose. This flexibility means there is no need to pay a premium for a so-called turnkey solution. Additionally, because VergeOS is licensed by the physical server, not the contents of that server, as you scale your environment, you can use quad-socket servers with massive core counts and not get crushed in licensing costs.

As we discuss in our article “The High Cost of Dedicated Storage”, VergeOS also dramatically lowers the cost of storage. You can use off-the-shelf server-class flash and hard disk drives, eliminating the 5X to 10X markup levied by dedicated storage array vendors.

The savings even follow through to the network. With VergeOS, you can use off-the-shelf commodity switches. When ready, you can replace dedicated network appliances like firewalls with VergeOS’ built-in L2 and L3 networking capabilities.

Conclusion

In conclusion, Hyperconverged Infrastructure (HCI) makes a poor foundation for an alternative to VMware, particularly in CPU performance, storage efficiency, scalability, and cost. Despite HCI’s initial promises of simplifying operations and reducing costs, it typically fails to meet these objectives due to its inherent inefficiencies.

Yet the cost and complexities of three-tier architectures remain. This has led to the development of VergeIO’s UltraConverged Infrastructure (UCI), which addresses HCI’s shortcomings as well as the three-tier challenges. VergeOS integrates infrastructure software modules into a single, cohesive codebase, enhancing performance and scalability while reducing costs. As such, it presents itself as a more viable solution for customers seeking a VMware alternative as well as for those seeking a simpler, more scalable data center infrastructure.

Learn More

Filed Under: HCI Tagged With: , ,

by

the impact of VMware's inefficiency

Licensing costs contribute to VMware’s total costs, but IT professionals almost always underestimate the impact of VMware’s inefficiency. They listed licensing costs as their primary concern in our recent survey. Still, other concerns, like required premature server and storage upgrades, lack of virtual machine density, and continual investments in backup infrastructure, highlight its inefficiency.

VMware’s inefficiency is brought on by years of plugging holes in the product through a never-ending series of bolt-on fixes, which often increases licensing costs and requires more hardware than necessary and more IT professionals to manage an increasingly complex environment. Even if VMware were to freeze its prices or even lower them, the ripple impact of its inefficiency makes a VMware exit to a more efficient alternative platform a wise strategy.

Understanding the Layers of Data Center Infrastructure

Most vendors divide the data center infrastructure into three layers. A hypervisor like VMware ESXi virtualizes the compute layer. A storage layer that in most medium to large-sized data centers is a dedicated storage array because of the shortcomings of virtualized storage products like VMware’s vSAN, and a networking layer that is built using proprietary networking hardware because of the expense of software-defined networking solutions like VMware’s NSX.

While it isn’t forgotten about, there is a fourth layer that is, for some reason, not included in the typical infrastructure discussion: backup and disaster recovery. In theory, with the right hypervisor software and storage capabilities, there should be no need for this to be a separate layer. Still, because of shortcomings, most organizations invest a large part of the IT budget into it.

The Impact of VMware’s Inefficient HCI

VMware started as a server virtualization and consolidation solution. Before VMware, each application ran on a dedicated server. VMware made it so IT could safely run multiple applications on a single server as virtual machines (VM). The idea behind hyperconverged infrastructure (HCI) is to extend that concept and eliminate the need for dedicated storage arrays and network appliances by moving the software that drives these dedicated appliances onto the same set of servers that VMware was using.

the impact of VMware's inefficiency

In the same way that VMware lowered costs by eliminating the need for a server for every application, HCI should lower costs by eliminating the need for a dedicated set of appliances for every aspect of storage and networking. The impact of VMware’s inefficiency means that most data centers have chosen not to use HCI and instead continue to use a legacy three-tier architecture. The problem is that most HCI vendors continue to use ESXi and run their storage or networking software as a virtual machine under ESXi, meaning they must navigate through the same VMware tax overhead as virtualized applications.

The result is HCI has not achieved the price advantages nor the operational simplicity that the original entrants into the market claimed. As a result, the legacy three-tier infrastructure is still the most common architecture in data centers.

HCI’s Inefficient Networking Problem

HCI also has an east-west inter-node network issue since storage and network operations must coordinate separately across all the cluster servers. Since each of these services is separate from the hypervisor, it triples the amount of inter-node communication, which limits scalability.

Because of the impact of VMware’s inefficiency, HCI networking limitations go beyond east-west traffic issues. In most cases, HCI vendors offer almost no additional network functionality besides what is embedded into the hypervisor. Or, in the case of VMware’s NSX, which is reasonably robust, it is not included as part of the hypervisor, and the add-on cost significantly increases the VMware license cost. In a recent blog, VMware suggested the best way to overcome its ransomware vulnerabilities was to deploy NSX, which, not coincidentally, almost triples the cost of the license.

VergeOS Eliminates the Layers

VergeIO is Ultraconverged Infrastructure; instead of re-creating the data center layers in software, VergeOS unifies them, including secure data protection, into a single code base that dramatically increases efficiency. Its efficiency lowers costs, enabling IT to do more with its existing resources while simplifying operations. Most customers can reduce physical server demand by 30% or more, which means running existing servers longer and ordering new servers less frequently.

VergeOS provides complete Layer 2 and Layer 3 network functionality, eliminating the concerns over east-west traffic contention and the need for separate network appliances. It also eases administration as network management becomes a seamless part of the data center infrastructure.

The Impact of VMware’s Inefficient Storage

Most HCI vendors are storage vendors in disguise. They make a storage software solution that can run as a VM within a hypervisor, typically VMware. These solutions, including vSAN, suffer from performance issues, partly because of vendors’ VMware performance tax and odd development choices.

Like HCI in general, storage, specifically in this architecture, should deliver rather significant cost savings and deliver better performance with the right architecture design. IT should be able to add capacity to their existing servers for a fraction of the cost of a “shelf upgrade” using a dedicated storage array. However, the impact of VMware’s inefficiency is felt most severely when HCI tries to provide storage performance and services comparable to a dedicated storage array.

In the survey above, we spoke to a customer looking to add about 300TBs of storage to their three-year-old All-Flash Array. HCI, if it had an efficient storage capability, should eliminate the AFA from consideration because the customer had several servers with twelve or more available storage expansion bays.

A 15.36TB NVMe SSD can be easily had for about $1,500. That means the customer can get 300TB of raw NVMe capacity, delivering well over one million IOPS for about $30,000, and insert them into the empty drive bays in their existing servers. When asked what it would cost to add 300TB of raw capacity to their Pure Storage array, they said at least 10X that cost.

As you can see, storage should be the area where HCI enjoys a significant price advantage, but again, like in other areas, it falls short:

  1. Since most HCI vendors are storage vendors, charging by the amount of capacity in use, the cost to add storage to HCI quickly rivals that of dedicated storage arrays.
  2. Most HCI solutions can’t add storage to available drive bays on just a few servers in the HCI instance. The capacity must be added to all the servers, or the customer must buy additional servers that match those in place.
  3. HCI inefficiencies mean the solutions can’t reach anything close to the performance potential of NVMe flash drives.
  4. Using storage services like deduplication, data protection, and snapshot retention times, further impacts storage and compute performance.

As an example of how storage services impact HCI’s efficiency, look no further than deduplication. While most HCI solutions support deduplication for capacity efficiency, in most cases, it is a bolt-on solution and not part of the original code base. VMware vSAN, for example, added deduplication years after it first appeared on the market. Nutanix’s deduplication appeared almost a decade after the product first shipped.

As a result, using deduplication from these vendors typically requires the addition of more powerful processors, additional memory, and, in some cases, changes to data protection strategies.

Limitations and compromises are the impact of implementing almost any storage feature as part of a virtualized environment:

  • Snapshots are limited to 32 total, and the best practice is not to have a snapshot that is older than a few days.
  • Data protection from media failure is complex and takes an undue toll on inter-node connectivity.
  • Replication for disaster recovery almost always requires a separate product.

VergeOS Supercharges HCI Storage Performance

Storage services are built into VergeOS and run as an equal citizen to the hypervisor instead of a subservient VM. Storage capacity can be inserted into nodes “post-facto” or storage-mostly nodes (a couple of processors and storage) can be added to the existing instance.

VergeOS provides a complete suite of storage services like deduplication, replication, and unlimited snapshots that can act as backups because they are more like clones, than traditional snapshot technology. Because storage is an equal citizen in VergeOS, our storage software is equally efficient and overcomes the challenges IT faces with VMware’s inefficient approach to storage. All these services can run without impacting performance, limiting their use, or forcing IT to make compromises elsewhere.

It is essential to highlight the VergeIO common sense licensing model. VergeOS is licensed by the physical server, not the number of processors, cores, amount of RAM, or storage capacity. In the above example, the customer can add 300TB of capacity with zero additional licensing charge.

The Impact of VMware’s Inefficient Data Protection

VMware also provides a very inefficient means of data protection, forcing all customers to create a separate backup and disaster recovery infrastructure. While it has basic protection from media failure and the ability, at an extra cost, to migrate VMs if a physical server fails, its snapshot capabilities are anemic at best. And their own best practices state, “Do not use VMware snapshots as backups.” For the most part, VMware’s snapshots are only used one at a time to provide data to a backup application and then quickly removed for fear of negative performance impact.

As a result, most customers implement separate backup software, which sends data to a separate backup storage area, which then must send it to another immutable backup storage area to protect against ransomware. These customers typically have a separate disaster recovery (DR) solution replicating data to a DR site. VMware’s inability to adequately protect itself is the cause of all of this additional investment in data protection.

The backup and recovery infrastructure becomes a separate cost and management point, often requiring specialized IT personnel. However, this additional investment does more than add to the total cost of VMware’s inefficiency. It complicates other tasks, such as patch upgrades.

For example, in our survey, we spoke to a VMware customer using HPE Zerto as a more powerful disaster recovery tool because of concerns about VMware’s ransomware vulnerabilities. There is an obvious cost concern with Zerto, but this customer’s current experience highlights a concern with the bolt-on approach caused by VMware’s inefficient approach to data protection.

In this case, the customer had a critical patch from VMware that closed a vulnerability in ESXi to a known ransomware exploit. However, they found that Zerto was not yet compatible with this latest release of VMware, and it would be at least three months before they were. The use of bolt-on technologies forces the customer into an awkward position. Do they deploy the VMware update and go without disaster recovery for three months, or do they keep disaster recovery working but put the organization at risk for a known ransomware exploit? While this situation does not increase the customer’s hard cost, the impact of VMware’s inefficiency certainly increases their mental overhead and anguish.

VergeOS is Secure and Resilient

Unlike VMware’s inefficient ransomware protection, VergeOS was built from the start to be a secure infrastructure software. The OS itself is hardened against attack. When combined with virtual data centers, immutable snapshots, and rapid alerting of encryption activity, customers can bounce back from ransomware threats in minutes with no data loss.

Conclusion

While licensing costs are a valid concern, the total cost of VMware’s inefficiency goes well beyond the surface. Inefficient HCI, storage, and data protection layers contribute significantly to operational complexity and expenses. VergeIO’s innovative approach with VergeOS offers a comprehensive solution that eliminates these inefficiencies, reducing costs, and simplifying operations. As data centers evolve, it’s essential to consider alternatives that optimize efficiency and empower IT professionals to do more with existing resources.

To learn more about how VergeOS can revolutionize your data center, watch our on-demand webinar as we discuss the results of our survey of almost 200 VMware customers and provide a live demonstration of VergeOS recovering from a ransomware attack.

Filed Under: VMwareExit Tagged With: , ,

by

Hyperconverged Infrastructures (HCI) were supposed to ease IT professionals’ management burdens, but flaws in their design explain why HCI doesn’t simplify IT. These flaws have led HCI, which should be the one infrastructure for the enterprise, to be a niche solution for corner use cases within the data center. Instead of simplifying through convergence, HCI solutions from companies like VMware and Nutanix exacerbate complexity.

The HCI Flaws

  1. HCI Isn’t Scalable
  2. HCI Isn’t Flexible
  3. HCI Isn’t Secure

Because of these three flaws, complex and expensive three-tier architectures with dedicated storage arrays, restrictive compute tiers, and proprietary network switches continue to flourish despite their own challenges. Can HCI be fixed so it can finally deliver its full promise?

HCI Isn’t Scalable

How can an architecture like HCI that is scale out by design, not scale? It’s true, though; HCI can’t scale large, and it can’t scale small. Most HCI vendors force you to start with three nodes, making HCI impractical for Edge and small offices. The inability to scale small and scale large is one reason why HCI doesn’t simplify IT.

Lack of a unified code base is one reason Why HCI doesn't simplify IT
HCI Doesn’t Converge, It Squeezes

At the same time, most HCI solutions can’t scale beyond eight or so nodes without serious performance concerns. These performance concerns stem from the fact that HCI solutions don’t actually converge the three data center tiers (networking, computing, and storage). They are three separate software packages, often from three separate vendors. HCI squeezes the three separate software-based tiers onto a single server. Each of these tiers has different lanes of communication between the servers. As a result, internode communication is exponentially increased by a factor of three.

UCI Solves the HCI Scale Issue

Ultraconverged Infrastructure (UCI) solutions, like VergeOS, solve both aspects of the scale issue. It can scale small, and it can scale large. Instead of using a separate application for networking, another for compute, and still another for storage, UCI is a single piece of very efficient code.

Lack of a unified code base is one reason HCI doesn't simplify IT. UCI has a Unified Code Base
UCI Converges Networking, Compute, and Storage Into a Single Code Base

Thanks to UCI’s unification of the legacy IT stack into a single piece of software, there is only a need for a single lane of communication between nodes. The reduction in east-west traffic enables scaling beyond 200 nodes without significant network overhead. At the same time, the efficiency of the code base enables a two-node UCI environment to deliver the performance and cost-effectiveness that Edge and remote locations need. The common code base also means a unified GUI for simple administration and day-to-day operations. UCI simplifies IT by providing a single software package. It can support small offices and Edge locations while scaling to meet the needs of the primary data center.

HCI Isn’t Flexible

Another reason why HCI doesn’t simplify IT is that it restricts the type of nodes users can add to the cluster. The idea behind HCI is that as you add additional servers to the environment, you automatically scale each of the three data center tiers to support new applications or additional users. The reality is that most organizations don’t need or want to scale all three of these tiers simultaneously; sometimes, you only need additional computing power, and other times, you only need additional storage capacity.

While some HCI solutions can now add “storage-only” nodes, they compromise performance and place restrictions on data protection and efficiency. Practically speaking, most HCI solutions must use nearly identical nodes as they scale, or create an entirely separate instance with an entirely different node configuration. Each additional HCI instance further adds to data center fragmentation instead of converging it.

In some cases, vendors require customers to replace servers, even if they are only a few years old, to support the latest version of their software. This requirement is not optional with them; upgrading to the new hardware means replacing older hardware.

UCI Solves the Flexibility Issue

UCI simplifies IT by solving the flexibility issue. It allows a mixture of different nodes within the same instance. Within the instance, VergeOS can group like nodes together. Then IT can allocate some or all those resources to specific workloads. This flexible intelligence also means that customers don’t have to upgrade or replace hardware to run the latest version of the HCI software.

The efficiency of VergeOS enables customers to extract more serviceable life from aging servers while fully exploiting the capabilities of modern servers that are only a few months old. Servers within a single VergeOS instance can be from different decades, processor manufacturers, and storage types.

Lack of a flexible scale is another reason why HCI doesn't simplify IT. UCI delivers complete flexibility.

Nodes can also have different areas of focus. Some can be heavily weighted for computing power while using the storage from other nodes in the instance, and others can be heavily weighted for storage capacity or performance and provide that storage or performance to the rest of the instance.

HCI Isn’t Secure

The lack of a robust security strategy is another reason why HCI doesn’t simplify IT. Customers must fill data protection and ransomware resiliency gaps by using third-party applications and creating additional infrastructures. As a result, legacy HCI forces IT planners to look at threats like ransomware as a backup and recovery problem instead of as an infrastructure problem, which is what it really is.

After IT discovers a ransomware attack, the malware file must be found, stopped, and removed. Then, IT must methodically recover data. By only looking at ransomware as a data protection problem, the average downtime associated with a ransomware attack is typically measured in days and sometimes weeks.

UCI is Hardened and Secure

UCI simplifies IT by providing infrastructure-wide data protection and resiliency capabilities. First, VergeIO’s Virtual Data Centers (VDC), which encapsulate the entire data center, also shrink the attack surface available to ransomware. It is nearly impossible for a malware file infecting one VDC to cross over to another VDC.

Second, VergeOS’ snapshot capabilities are powered by IOclone, which provides independent, immutable copies. However, because VergeOS implements global inline deduplication at its core, clone/snapshot copies are also space efficient. Taking frequent immutable snapshots without impacting performance provides the second layer of defense against ransomware.

Third, VergeOS’ IOfortify provides early detection of a ransomware attack. With it, your notification comes minutes after an attack starts instead of coming from confused users hours later. This early detection, combined with frequent, immutable clones, means recovery occurs within minutes and with little to no data loss.

Finally, VergeOS operates as read-only firmware at its core and injects a copy of itself into each VDC. As a result, if a cyber-attack could ever get to the actual VergeOS operating environment, a known good copy at the core is available to replace it.

The Impact of HCI Complexity

The impact of HCI complexity and its limitations means that IT must use the legacy three-tier architecture. It is also why most IT professionals don’t think HCI can replace a SAN, even though it should be a better option for virtualized environments. While legacy three-tier architectures are more complicated and more expensive, they, through brute force, work through some of the issues mentioned above.

UCI fulfills and extends the potential of legacy HCI. Using a single code base allows IT to manage all aspects of infrastructure in a cohesive, straightforward manner. UCI has flexible scalability, extending to both large and small needs. UCI can mix nodes of different types, which enables true and long-lasting actual convergence. UCI’s integrated data protection capabilities eliminate the dependency on a separate data protection infrastructure. As a result, UCI delivers more widespread convergence than HCI, while delivering infrastructure-wide data resiliency.

To learn more about UCI and VergeIO, register to watch the VergeOS Architecture Deep Dive with our CTO and founder, Greg Campbell. You can also sign up for a virtual test drive and start running VergeOS in minutes.

Filed Under: HCI Tagged With: ,

by

IT has to factor the cost of the VMware Tax into its infrastructure planning. Where does this tax come from? It is the extra expense in hardware needed, to compensate for the overhead of VMware’s inefficient virtualization code. Adding additional components like vSAN for storage, or NSX for networking only makes the tax more severe. These components will adversely impact applications and users if IT doesn’t work around the overhead. These workarounds cost money, increase complexity and create a more brittle infrastructure that struggles to adapt to the organization’s future needs.

There are three primary effects of the VMware Tax:

  • A lower-than-possible VM Density.
  • The continued need for stand-alone bare-metal workloads.
  • A proliferation of the three-tier architecture.

To compensate, organizations are forced to:

  • Buy more or more powerful physical servers than what should be necessary.
  • Buy high-performance dedicated all-flash arrays.
  • Face a never-ending future of premature server and storage refreshes.

The Cost of the VMware Tax is one of the hidden costs of VMware. Learn about the rest of them in our on-demand webinar, “The 4 Hidden Costs of VMware.”

The Cost of VMware Tax on Server Planning

The first cost of the VMware Tax is that meeting the organization’s demands requires using either fewer virtual machines (VMs) per server and more physical servers or more powerful physical servers to support more VMs per server. As we discuss in our article, “HCI isn’t Infrastructure,” using more servers highlights the scaling issues common in Hyperconverged Infrastructure (HCI) and leads most customers to purchase more powerful servers to reduce cluster node count.

Using more powerful servers with more CPUs, cores, and RAM increases those servers’ costs. This approach also increases VMware licensing costs since the company charges by the physical CPU installed in its nodes, and the expectation is that after the Broadcom acquisition is complete, the company will switch to a per-core licensing strategy and is expected to raise costs even further.

Overcoming the VMware Tax with Powerful Servers

Overcoming the cost of the VMware Tax by using more powerful servers exposes the lack of flexibility when scaling, the typical VMware architecture has. More powerful servers will likely be used longer before organizational growth requires IT to add another server to the cluster. When it comes time to add additional servers to the cluster, two or three years later, that exact server type may no longer be available, or a better option may be available using a different CPU vendor. Managing nodes of different types within a VMware environment is convoluted, leading many customers to establish an entirely separate cluster, further increasing costs and complexity.

There is also the challenge of, after three or four years, the hypervisor or storage vendor may upgrade its software to the point that it no longer supports the original servers. At that point, IT is faced with refreshing its entire server infrastructure to maintain compatibility with the updated software.

The Cost of the VMware Tax on Scale

Hyperconverged Infrastructure (HCI) solutions expect a “balanced scale,” where the organization adds equal amounts of computing, storage, and networking. This requirement exposes another cost of the VMware Tax. No organization always needs to scale these three components at the same time. Especially when purchasing more powerful servers, the organization will likely need more storage performance or capacity before requiring additional computing resources.

Again, legacy HCI designs like those from VMware (vSAN, vSphere, NSX) or Nutanix expect almost identical nodes to be added to the cluster. To meet capacity demand, customers are adding servers similar to their original installation, and they end up paying for and wasting a massive amount of CPU and memory that come with those servers. Although a few vendors have evolved to allow more storage-centric nodes, these are complex band-aids with many compromises and increasing complexity. Most customers choose not to use them.

The enforcement of balanced scaling forces most customers considering HCI solutions to disqualify them later. As a result, the VMware Tax indirectly proliferates the more complex and expensive three-tier architecture.

VergeOS Enables High VM Density and Imbalanced Scale

VergeOS is a highly efficient data center operating system (DCOS) that enables high per-physical server VM populations. It does this by integrating the three standard data center tiers (networking, computing, and storage) into a single code base that is a fraction of the size in terms of lines of code without sacrificing features. This effort is the foundation of Ultraconverged Infrastructure (UCI) which moves beyond the flawed HCI model to deliver the full promise of a truly converged infrastructure.

A more compact code base means faster execution on the same hardware. Integrating the traditional data center tiers means that efficiency carries throughout the entire solution. Our typical customer can support 25 to 30% more VMs per physical server with VergeOS than their prior solution (Hyper-V, VMware, Nutanix) while running on the very same, existing hardware.

The lack of flexible scale increases The Cost of the VMware Tax

As the organization’s demands grow, VergeOS provides flexible and intelligent scaling. An optimized internal-node communication protocol ensures near-linear performance increments as IT adds nodes to the environment. VergeOS’ flexibility means that IT can add nodes of almost any type, including storage or compute-centric nodes. VergeIO has never forced the retirement of old servers to support a new version of VergeOS.

VMware Tax Encourages Bare Metal

Despite all the advantages and flexibility of virtualization, the cost of the VMware Tax means that many customers still view some applications as bare-metal only. The performance demands of these servers are just too strenuous for the virtual environment. They either starve other VMs on the same node of resources or can’t continually guarantee access to the performance these applications demand. They also tend to be certain hardware types, like GPUs, that specific applications need, which legacy solutions don’t adequately virtualize.

The result is IT must purchase and stand up dedicated silos of computing and storage for these applications, which increases costs and complexity. Additionally, these bare-metal environments don’t benefit from core virtualization capabilities like seamless VM migration.

VergeOS Delivers Near Bare Metal Performance

VergeOS’ compute efficiency also helps formerly bare-metal-only workloads to be able to realize the benefits of virtualization. Historically, many bare-metal workloads remained bare-metal because of the storage I/O demands. VergeOS integration of storage, coupled with the performance of its file system, delivers near-bare-metal performance for these workloads. The VergeOS storage component, VergeFS, ensures all reads are local to the virtual machine. Finally, VergeOS supports nodes of various CPU classes within the same instance. Our customers can and do mix Intel, AMD, and GPU-based nodes, even of different generations. They repeatedly tell us that they were able to virtualize formerly bare-metal workloads without any degradation of performance.

Conclusion

With its inefficiencies and subsequent costs, the VMware Tax represents a significant concern for IT infrastructure planning. Its impact on VM density and the push toward bare-metal workloads signifies the increased expenses and complexities organizations face. As the technology landscape continues evolving, solutions like VergeOS emerge as viable alternatives, offering enhanced VM densities and near bare-metal performance.

VMware VM to VergeOS Chalktalk

By merging the conventional data center tiers and focusing on efficient operation, VergeOS addresses the challenges posed by VMware’s model and presents a forward-thinking solution that adapts to an organization’s dynamic needs. As organizations strive for agility, cost-effectiveness, and scalability, transitioning to platforms prioritizing these attributes will be paramount. VergeOS makes that transition seamless thanks to our IOmigrate capability.

Filed Under: VMwareExit Tagged With: ,

by

HCI isn't an Infrastructure

As IT leaders consider the practicality of a VMware exit, they will also consider hyperconverged infrastructure. However, the inability to scale flexibly means that HCI isn’t an infrastructure they can use as a replacement platform. Even the few HCI solutions that are not dependent on VMware as their hypervisor suffer the problem of brittle scalability.

The Three Requirements of Flexible Infrastructure

A flexible infrastructure must meet these three requirements:

  1. The ability to scale small, less than three nodes for Edge, Remote Office, and small business data centers.
  2. The ability to scale large, dozens to hundreds of nodes, to meet the demands of an enterprise data center.
  3. The ability to manage nodes of different types, CPU brands, compute-only, and storage-only, so that it can adapt to the ever-changing needs of the data center.

The inability to meet these requirements with a single solution means that if an organization selects HCI as its infrastructure strategy, it likely will either need multiple HCI solutions to cover the organization’s needs, or at least multiple instances of the same HCI solution. Ironically, Hyperconverged Infrastructure (HCI) isn’t an infrastructure, nor does it converge.

UCI is Infrastructure

Because of its inflexibility, a counter to HCI, which is relegated to niche use cases, is Ultraconverged Infrastructure (UCI). UCI is an infrastructure, Data Center Operating System (DCOS), and does truly provide convergence. VergeOS is a UCI solution that can scale small, large, and is flexible which allows it to adapt to the changing needs of organizations. Additionally, it is not dependent on VMware, which makes it an ideal alternative for customers looking for a VMware exit.

Infrastructure Must Scale Small

Not every business is an enterprise, but most need compute capabilities beyond what cloud-based SaaS applications can provide. As those businesses scale, the cost of the cloud becomes a significant factor. The point of entry for most HCI solutions is a three-server configuration which becomes nodes in its cluster. This requirement is often too large for Edge, Remote Office, and Small Data Center use cases.

As a result of the inability to scale small, most IT planners rule out HCI for these use cases and compromise with standalone servers and an inexpensive shared storage solution. The compromise increases the cost of acquiring the hardware and software for Edge, Remote Office, and Small Data Center use cases, and it increases complexity in an area that must have simplicity. These use cases have limited or no IT administrators, especially at the Edge.

UCI Scales Small

Users can start with as few as two nodes, and because of VergeOS’ efficiency, those nodes can be very cost-effective mini-servers, ideal for the Edge or Remote Office use cases. Even small data centers can run comfortably on two or three low-end to mid-range servers, reducing hardware acquisition costs by 50% or more. Learn more about the advantages of using UCI in small data centers here.

VergeOS’ efficiency comes from its tight integration of the network, compute, and storage tiers into a single, cohesive data center operating system (DCOS). The actual convergence of those tiers, instead of HCI’s approach of an “elegant bundle,” means the elimination of redundant metadata tables, databases, APIs, and management tools. The result is that the underlying hardware is unchained and enabled to reach its full potential.

Infrastructures Must Scale Large

Small businesses grow to medium-sized businesses and then eventually to large businesses. Starting over with a new infrastructure as the business grows is costly and risky.

Again, most HCI solutions start too big to be viable for small data centers or Edge, but the few that focus on the smaller end of the market often can’t scale large enough to meet the organization’s demands as it grows. They often can’t scale past six to eight nodes. The result is IT must replace the “starter-HCI” solution with another more scalable solution. The replacement often includes replacing hardware since each HCI solution seems to have a unique hardware compatibility list or comes “bundled” with hardware.

Even so-called enterprise HCI solutions have limitations and can scale to only a few dozen nodes, which is not large enough to provide complete infrastructure consolidation. There is also a practical limit to how large HCI solutions can scale. Most vendors use standard IO protocols to communicate between nodes, which means that every node must be “touched” every time a packet is received by the cluster, creating an untenable amount of internode, or east-west traffic.

Yet another part of the challenge is that most HCI solutions don’t include complete layer 2 and layer 3 networking functionality in their product. At most, they provide virtual switching or they bundle in, yet again, another third-party software-defined networking product. The lack of native networking functionality means that HCI clusters can’t practically scale to more than a dozen nodes without suffering a performance impact.

These challenges are why the legacy three-tier architecture continues to be the bread-and-butter infrastructure for enterprises. It is also why most IT leaders believe that HCI cannot replace standalone networking, virtualization, and storage.

HCI isn't an Infrastructure

UCI Scales Large

VergeOS can not only start as small as two nodes, but it can grow to well over one hundred nodes. Organizations of any size can start using VergeOS with confidence that it can grow or shrink, to meet their needs. The team at VergeIO also developed a proprietary networking protocol that optimizes internode communication, significantly reducing east-west traffic and making scale technically possible and practical.

Infrastructure Must Scale Flexibly

Infrastructure flexibility is a critical requirement because small and large businesses evolve. Their IT needs are not static, and neither should their infrastructure be. It needs to adapt to innovations in hardware and business needs.

Most HCI solutions are rigid in their configuration, only supporting specific hardware and forcing customers to upgrade to new hardware as the infrastructure software is updated. In HCI terms, scaling up typically means adding identical nodes, making it difficult to take advantage of the latest hardware advancements or adopt new types of nodes, such as storage-only or compute-only. HCI’s lack of flexible scaling also poses challenges when organizations want to retire or replace outdated equipment. Mixing old and new equipment within the same cluster is almost impossible.

The typical HCI setup often means organizations must create separate instances for workloads and hardware types. This approach fragments the overall infrastructure and creates more complexity, reducing operational efficiency and leading to underutilized resources. The inability to mix and match different types of nodes within the same cluster further diminishes the flexibility and cost-effectiveness of HCI.

UCI Scales Flexibly

On the other hand, UCI provides an answer to the rigidness of HCI. VergeOS supports multiple types of nodes, such as compute-only or storage-only nodes, which allows it to scale flexibly to match the dynamic needs of the business. As technology evolves and new hardware becomes available, VergeOS users can integrate these advances seamlessly into their existing setup. Customers can, for example, mix in AMD, Intel, and GPU nodes into the same instance.

Furthermore, with VergeOS, it’s possible to maintain a diverse set of hardware in a single instance, preventing infrastructure fragmentation. This capability allows businesses to adjust their IT setup as the organization grows and its needs change, ensuring they always have the most cost-effective and efficient infrastructure.

HCI isn't an Infrastructure


Conclusion

VergeOS is an ideal solution for businesses seeking a flexible, scalable, and efficient IT infrastructure. UCI surpasses HCI in meeting the demands of a dynamic business environment, ensuring that businesses can focus on their core competencies without worrying about their infrastructure. Whether it’s a small enterprise looking to grow or a large organization needing to maintain agility and efficiency, VergeOS has the features and flexibility to accommodate their needs. Compare HCI with UCI here.

To learn more about scaling IT infrastructure, watch our on-demand webinar, “How to Eliminate the Data Center Scale Problem.”

Filed Under: HCI Tagged With: ,