CPU and memory performance has increased geometrically while storage performance
has lagged behind. RAID arrays have improved to help reduce this
gap but can suffer from slow performance due to excessive I/O. Slow RAID array performance
can bottleneck a whole system slowing your applications.
The Windows OS inherently splits files apart and scatters them around a volume instead
of writing them in one place contiguously. When this happens a large number of unnecessary
requests are sent to the RAID controller.
This is where V-locity I/O optimization software
comes in—by proactively preventing unnecessary I/O at the Windows OS level. Preventing
unnecessary I/O not only prevents the degradation in VM and application performance,
but improves RAID array performance, keeping it running at peak speeds.
Get IDC's Technology Spotlight: How to Manage the I/O Explosion with
No Additional Hardware »
Solution to Increasing RAID Array Performance
A file system generates a lot of unnecessary I/Os, however, the RAID controller
is unaware that the multiple I/Os are part of the same file and treats each I/O
as a separate entity. This leads to RAID array performance degradation. Preventing
excessive and unnecessary I/O at the file system level and consolidating data into
a single I/O improves the performance of write and read activities.
V-locity® I/O optimization software
uniquely addresses the unnecessary I/O patterns of reads and writes by aggregating
data on writes to perform sequentially while caching reads on available server memory
without contention to the application. As a result, V-locity boosts application
performance by 50% or more by (1) maximizing the efficiency of every single I/O
from write requests that leave the server and (2) caching the most active data from
read requests using available server memory.
Read I/O Optimization – Much of the I/O traffic moving through
the infrastructure shares similar data, yet every byte of that data travels the
entire distance from server to storage and back for every I/O request, even when
it is frequently accessed data. IntelliMemory™ technology addresses this read request
inefficiency by caching the most active data in the server's available memory. Should
an application require more memory, V-locity serves memory back to the application
to ensure there is never a case of memory starvation or resource contention. Condusiv
Technologies' self-learning caching algorithms have set the gold standard for the
industry as they are OEM'd by eight of the top ten largest PC manufacturers in the
world.
Write I/O Optimization – V-locity's IntelliWrite®
technology optimizes I/O write operations, eliminating the performance penalties
associated with the Windows OS, splitting files into pieces and writing each piece
to a different logical location within the SAN or NAS. V-locity is aware of space
allocation and aggregates writes to behave sequentially—requiring less I/O for every
file written. Subsequent reads also benefit, since only minimum I/O is required
to fulfill the request. As a result, only productive I/O requests are processed
across the entire infrastructure, enabling more data to be processed in the same
amount of time.
When only productive I/O traffic is processed through the server, network and storage,
I/Os per second are accelerated, latency is greatly reduced, and more work can be
performed in the same amount of time. This greatly improves the efficiency of all
VMware ESX/ESXi and Microsoft Hyper-V virtual platforms and physical servers for
increased bandwidth.
V-locity I/O optimization is tailored for virtual environments that leverage a SAN
or NAS as it proactively provides I/O benefit—improving storage performance and
benefiting advanced storage features like snapshots, replication, data deduplication
and thin provisioning. Since V-locity optimizes I/O at the server level, it is complementary
to all SAN and NAS storage systems and media types (e.g., HDD or SSD).
Download a trial of V-locity and put it to the test in your real-world
environment »