By Vinod Mohan of DataCore SoftwareTransform data into insight and information into intelligence with the power of analytics and machine learning.
Everything
runs on data in today's digital economy. In this data-driven world,
ensuring your business data is securely stored, highly available, and
rapidly accessible on your storage devices is mission critical. As with
any IT hardware, storage infrastructure is also prone to failure,
performance slowdowns, and capacity exhaustion, each of which adversely
impacts data processing in the back-end and disrupts business continuity
in the front-end.
It
is the responsibility of any IT organization to efficiently manage
their storage infrastructure, adequately plan for capacity expansion,
and implement robust data protection and business continuity practices. This is where storage analytics comes to your aid.
Metamorphosis of Information into Intelligence
There
is a wealth of information contained in the KPIs obtained from your
storage infrastructure, such as IOPS, bandwidth, faults, latencies, etc.
which help understand health and performance status. Operational data
such as capacity allocation and usage, device configuration, and
inventory details provide additional context about the storage devices.
Collecting
and analyzing this infrastructure data over time can reveal resource
usage trends, anomalies, and patterns of failures which help forewarn of
impending issues and preemptively avert them before any performance
impact. Machine learning techniques can process all this data and
transform the information contained in them into intelligence. Storage
analytics tools incorporate AI/ML algorithms to process historical and
current data to forecast trends and predict problems. The effectiveness
of analytics typically depends on the volume and variety of data
collected, frequency of telemetry, and retention period.
In this blog, we will look at four different types of analytics that are available with DataCore Insight Services (DIS),
a purpose-built data storage analytics solution that helps IT teams
gain actionable intelligence to troubleshoot storage bottlenecks,
preclude performance-impacting problems and improve decision-making on
capacity planning and storage expansion. Let's dive in and understand
the importance of all four types of analytics.
#1 Descriptive Analytics
This
is the simplest form of analytics where chunks of aggregated data are
processed and condensed into useful nuggets of information which are
descriptive of an insight. From the combination of historical and
real-time data gathered from the virtual storage infrastructure managed
by DataCore SANsymphony software-defined storage solution,
DIS provides many descriptive insights. For example, whether a virtual
disk has failed, a latency threshold has been exceeded, a disk pool is
running out of capacity, remote replication is stuck, a snapshot has
failed, there is link error on a server port, etc. Alerts can be set up
to notify of issues happening in real-time. For a given time period in
history for which data is retained, analytics can also be run to
highlight issues that happened in the past.
In
the example below, DIS uses the power of descriptive analytics to rank
the most active hosts and vDisks in your storage infrastructure to help
you understand the I/O impact on your busy devices. As you can see,
individual sensors collected from different hosts and disks have been
relationally analyzed to present this meaningful and actionable
information.
Descriptive analytics transforming raw data into meaningful insight
#2 Predictive Analytics
This
is the need of the hour for all IT teams: the ability to know in
advance an issue is going to happen so that it can be prevented without
impacting data access and storage performance. Based on the capacity
utilization metrics collected from SANsymphony nodes over time, DIS
leverages its machine learning capabilities to highlight patterns and
extrapolate the numbers based on deterministic factors to a future
timeline in order to deliver forecasts. This is very helpful to answer
capacity-related questions and subsequent decision-making that storage
administrators are faced with all the time.
Based on the current workload:
- When will my storage pool run out of space?
- When should I add additional capacity to my storage pool?
- When should I start budgeting for new hardware purchase?
- Is it time to reclaim unused storage capacity from other available disks to meet the depleting space on my current device?
Visualize
predictive analytics on intuitive trend graphs that easily delineate
current usage pattern and expected growth in the future.
Typically,
countless hours are consumed in manual data collection, guesstimating
growth, and presenting a forecast. All this effort can be saved by
automating predictions with the help of storage analytics.
#3 Prescriptive Analytics
Prescriptive
analytics is the next stage of descriptive and predictive analytics
wherein based on intelligence gained, the analytics tool presents
suggestions and recommendations to fix the problem by taking
corrective/preventive measures. DIS incorporates the collective
knowledge of the community of DataCore users around the globe. Based on
years of experience in encountering storage issues such as failures,
configuration anomalies, resource contention, etc., DIS includes a
myriad of best practice recommendations provided in context of insights.
Storage administrators can follow these steps to resolve the
problem/alert that they are running into.
Example 1: When
running out of storage space in the storage pool, DIS would give you
recommendations to proactively optimize existing capacity for maximum
utilization.
Example 2: When
encountering a high I/O latency issue, DIS uses descriptive analytics
to present the details of the issue and then goes on to provide
prescriptive analytics to allow the storage administrator to resolve the
issue.
This
also helps reduce your outreach for support requests to either DataCore
or your storage vendor as you are now empowered with actionable
analytics to resolve commonly encountered storage-related issues.
#4 Diagnostic Analytics
At
the heart of troubleshooting is diagnosis. Only when a problem is
diagnosed properly and root cause identified, can it be triaged
effectively. Diagnostic analytics takes a deeper look at data to
understand the root cause of the events and is helpful in determining
what factors and events contributed to the outcome.
DIS includes analytics to help correlate problem patterns with anomalous events to help locate the cause of the problem.
- Is storage performance slowdown related to high I/O activity or is the storage controller not configured properly?
- Why
is capacity running out soon on primary storage? Are there unserved
virtual disks consuming space? Is there inactive (or infrequently
accessed) data getting stored on the device tying up space better suited
for active data?
- Why is synchronous mirror failing: Is it because of a inter-campus link failure?
The example below shows DIS analyzing whether disk pools are optimally configured across tiers. Especially during data tiering,
it is important to provision capacity to storage tiers in accordance
with how much hot/warm/cold data is getting stored on the devices. The
performance summary in the pattern below shows that hotter data is
getting stored on tier 2 (not the fastest storage) because tier 1
(fastest storage) seems to be fully allocated. Diagnosis by DIS reveals
this as the reason for sub-optimal performance. The prescriptive part of
the analytics (suggested actions) provides the recommendation to add
more storage capacity to tier 1 to ensure all the host data gets placed
on and accessed from it.
Intuitive storage pool performance diagnostic analytics in DIS
The Devil is in the Details: Analytics,
in conjunction with machine learning and AI capabilities, can
revolutionize how you plan, implement, and manage enterprise storage.
With the right insights and intelligence by your side, you can be
well-informed to improve capacity utilization, eliminate performance
bottlenecks and optimize workload placement. Ask your DataCore partner about DataCore Insight Services (DIS) today!
##
About the Author
Vinod Mohan is a Senior Product Marketing Manager at DataCore
Software. He has over a decade of experience in product, technology and
solution marketing of IT software and services spanning application
performance management, network, systems, virtualization, storage, IT
security and IT service management (ITSM). In his current capacity at
DataCore, Vinod focuses on communicating the value proposition of
software-defined storage to IT teams helping them benefit from
infrastructure cost savings, storage efficiency, performance
acceleration, and ultimate flexibility for storing and managing data.
Prior
to DataCore, Vinod held product marketing positions at eG Innovations
and SolarWinds, focusing on IT performance monitoring solutions. An avid
technology enthusiast, he is a contributing author to many popular
sites including APMdigest, VMblog, Cyber Defense Magazine, Citrix Blog,
The Hacker News, NetworkDataPedia, IT Briefcase, IT Pro Portal, and
more.