Hyperconvergence Facts and FAQs

© Springpath, Inc. 2015 All rights reserved | www.springpathinc.com| 02092015 1001
A TECHTARGET WHITE PAPER
Hyperconvergence Facts and FAQs
What you need to know about the next-generation
data center infrastructure for optimizing
VM-centric environments

1
1. Hyperconvergence simplifies the
deployment of highly available server
and storage clusters by working with a
data center’s existing server and storage
resources.
2.By eliminating silos between servers
and storage, hyperconvergence reduces
operational complexity and simplifies
data center, application and resource
management.
3.It is easy to add more resources to a
cluster without downtime, whether it’s
more compute or additional capacity.
What’s more, storage is never orphaned
in “islands of underutilization.”
4.Hyperconvergence is built on a
distributed file system that is purpose-
built and optimized for virtual machine
environments.
5. Data management capabilities—
snapshots and clones—are built
into the distributed file system, and
run transparently without affecting
performance.
TOP10
BENEFITS
© Springpath, Inc. 2015 All rights reserved | www.springpathinc.com
Next-Generation Data
Center Infrastructure with
Hyperconvergence
Cutting-edge data centers are most efficient when
built on a next-generation architectural foundation.
Take virtual machines. While today VMs are often
deployed on traditional architectures with servers
and separate network attached storage arrays,
enterprises considering broad use of VMs will see
the best results by implementing a new architecture
called hyperconvergence.
Hyperconvergence eliminates the silos that traditionally
separate a data center’s compute resources (servers)
from its storage resources (such as storage area
networks or network-attached storage).
In a virtualized environment, silos create operational
complexity, as you need coordination between servers
and storage teams. What’s more, multiple storage
arrays may have been purchased for specific use
cases, so admins need to manage each type of storage
separately. In an organization using storage-area
networks (SANs), there is often a LUN-to-VM construct
mismatch that can be difficult to overcome.
Silos also waste expensive storage space, underutilize
processors and memory, and can make it difficult to
reallocate or scale in cost-effective, agile manner.
Enter hyperconvergence, which integrates storage
with compute resources to provide a more efficient
data center with higher resource utilization, better
operational efficiency, easier management, and far
more flexible scalability.

© Springpath, Inc. 2015 All rights reserved | www.springpathinc.com 2
How it Works
Hyperconvergence creates one big pool of compute
and storage resources, which is ideal for virtual
machines. Clustering storage together with groups
of servers simplifies deployment of virtual machines
and allows for rapid allocation of storage resources.
Hyperconvergence is enabled through the use of a
purpose-built distributed file system that sits between
the hypervisor and all of the storage resources.
Indeed, hyperconvergence reduces operational
complexity and works transparently to the applications.
The distributed file system is invisible, providing a pool
of fast, scalable resources to virtual machines. There’s
no performance impact, no new tools to learn, no new
APIs to manage.
Hyperconvergence is designed for flexible, agile
scalability, no matter what the requirements. Data
center operators can add more storage, more compute
capacity, or both, wherever it is needed most. In fact,
with the right solution, operators can independently
scale data caching and data persistence tiers
independently. The best hyperconvergence systems
eliminate the wasted storage found in traditional silo-
based data centers to ensure that spinning hard disk
drives (HHDs) and solid state drives (SSDs) are fully
utilized. What’s more, deduplication and compression
are designed right into the distributed file system, and
are leveraged across the cluster.
Let’s learn more.
6. Data optimization capabilities—
global inline deduplication and inline
compression—are also built in, and
work automatically to improve storage
efficiency
7.Operators can improve performance
and capacity by scaling the caching tier
independently from the persistent tier,
as needed.
8.Hyperconvergence is self-healing;
data is mirrored to multiple disks and a
rebalancing engine keeps applications
running while failed
hardware is replaced
9.Software-based hyperconvergence
is easy to implement and easy to learn,
and administrators can manage it at
the VM level using tools like VMware’s
vCenter.
10. The economics of software-
based hyperconvergence can be
transformative, with low-cost, low-risk
annual licensing that lets data centers
experiment with this new technology.
TOP10
BENEFITScont.

Where does the hyperconvergence management layer go?
With hyperconvergence software, a distributed file system is installed onto the vSphere ESXi
server and users operate the hyperconvergence layer from vCenter. Installing and configuring
hyperconvergence on VMware takes less than 30 minutes to complete. With some vendors’ hardware-
based solutions, there is a separate management console that users must learn in order to allocate
and manage hyperconverged resources.
Hardware or software?
Software-only solutions are the best way to go. A software-based hyperconvergence solution allows you
to continue using your preferred server vendor, instead of introducing another server vendor into the
data center. And because it’s software, you can scale the specific elements you’d like—compute, hard
drives or SSDs—with no need to add entire nodes. From an operational standpoint, purchasing the
same servers that you have been buying will allow you to leverage existing server management tools,
such as Cisco UCS Manager or Dell OpenManage, and keep your server management environment
unified. When it’s time to refresh hardware, you can simply move the hyperconvergence software
licenses to the new servers—there is no need to buy expensive hyperconverged appliances.
What’s adaptive scaling?
A challenge with traditional data center architectures is that once a bottleneck appears or a storage
limit is reached, it can be very hard to scale to address that specific problem. With hyperconvergence,
adaptive scaling means that you can continuously scale out capacity, and specifically add only what is
needed, such as more compute-only blades, more HHDs or more SSDs to independently address the
tier that needs expansion.
How does hyperconvergence help optimize data and I/O?
Hyperconvergence is implemented as a distributed file system—and that unleashes powerful,
transparent capabilities across the cluster’s applications. For example, it has built-in capabilities for
inline deduplication, which can reduce storage utilization tremendously. Data compression is also built
in, not only saving storage but also reducing storage I/O requirements.
FREQUENTLY
ASKEDQUESTIONS

What about VM snapshots and clones?
It’s important to look for a hyperconvergence solution that provides fast and efficient snapshots and
clones of virtual machines. When implemented correctly, clones should be taken within seconds,
allowing for rapid deployment of multiple virtual machines for test/dev or production environments. In
addition, look for solutions that store only the differences once the snapshots/clones are created. The
outcome is very space-efficient copies, using the least amount of storage necessary.
What about resiliency?
With hyperconvergence, the distributed file system is self-healing. It pools all the storage resources in
a cluster, and automatically mirrors all data to multiple physical HHDs and SSDs. Writes are cached
and persisted in multiple locations until committed to multiple storage devices. If a storage unit fails,
the file system automatically serves data from the mirrors. The system also proactively monitors and
signals administrators to replace the faulty devices. Once replacement storage goes online, the file
system automatically redistributes the data and restores redundancy.
What is the Springpath hyperconvergence solution?
The Springpath Data Platform is purpose-built for hyperconvergence and integrates into VMware
and other infrastructure platforms to turn standard servers into a single pool of compute and storage
resources. Springpath Data Platform eliminates the need for network storage and using adaptive
scaling capabilities allows customers to grow compute, caching or capacity resources independently,
depending on their changing business needs. Springpath proactively monitors customers’
environment to ensure resilient, always-on availability. Using an advanced set of data management
and optimization capabilities such as compression, deduplication, space-efficient snapshots and
clones, users experience transformative levels of resource utilization at much lower TCO than
traditional IT architecture.
How can I learn more about the Springpath Data Platform?
Visit : http://springpathinc.com/smart-hyperconvergence.php
FREQUENTLY
ASKEDQUESTIONS

Hyperconvergence Facts and FAQs

More Related Content

What's hot (20)

Similar to Hyperconvergence Facts and FAQs (20)

Recently uploaded (20)

Hyperconvergence Facts and FAQs