Netreo whitepaper 5 ways to avoid it management becoming shelfware

To Keep Your
5 WAYS
IT MANAGEMENT
SOFTWARE
From Becoming
SHELFWARE
An Industry Perspective
April 2015

A leading industry analyst recently estimated that 84% of companies were paying for software
licenses that were either not fully deployed or had been neglected and marginalized to the point
of becoming‘shelfware’. This boneyard of shelfware products, idle and irrelevant, sits in the back
gathering dust and gobbling up budgets with its outdated lifecycle footprint before it’s finally put
to pasture. This process can go on for years.
It’s all too common to see Management software at the top of the heap for shelfware. We've been
able to identify some of the leading causes that prompt companies to abandon these (often
significant) software investments.
Here are 5 key pointers you can use to keep your IT management software investment producing
positive returns.
The Solution 6
We've been able to
identify some of
84%of companies are paying for
SHELFWARE
2
the leading causes
#1. Too many alerts are worse than not enough
#2. Don't make people go to multiple places to gather information
#3. Keep the interface simple and fast
#4. Minimize administration
#5. Don't buy software
3
3
4
4
5
page

Many companies, excited by the possibilities
inherent in their new management software,
are eager to configure every possible monitor,
alert, and threshold so that they can get maxi-
mum visibility into their environment. When
they do this, the resulting flood of alerts tends
to cause 'alert overload’among the IT staff, and
very quickly the alerts get ignored or filtered
out. If 90% of the alerts are 'not actionable' -
meaning that the person receiving the alert is
not expected to do anything about it - the
natural instinct is to overlook them. The real
challenge here is that when a critical issue
occurs, it is very often lost in the noise and
chaos of all the false alarms that are being
generated. It can also cause a false sense of
security, with users thinking that every possible
occurrence will be detected and reacted to. In
reality, we've seen many companies with this
kind of configuration still relying on the users
calling the help desk as their first sign that
something is really, seriously wrong.
The key thing to remember when configuring
alerts is to make sure that they are "actionable"
alerts - that the person receiving them is
intended to take immediate action to correct
the problem. Everything else can be handled
via automated reports to send the information
out on a daily, weekly, or even monthly basis.
The ultimate goal is to be proactive in address-
ing the conditions that cause them.
1. Too Many Alerts Are Worse Than Not Enough
We often find companies have allowed each
department to select and purchase their own
tool. This makes perfect sense in some cases,
like allowing the DBA team to implement
something very specialized for their specific
database monitoring needs. Debugging tools
for an in-house software development team
might be another example. But if left
unchecked, multiple departments using their
own tools can lead to lots of finger pointing.
Often when a problem occurs, none of the
individual 'silo' tools have spotted the issue,
and the departments each run around check-
ing their own areas of responsibility looking for
something that went wrong. This wastes a lot
of time and resources, and puts you in the
position where you're still relying on users as
your primary failure detection system.
In order to achieve real proactive management
and stay ahead of the curve, you need an
holistic view into all the different aspects of IT -
from the applications themselves, to the
servers delivering them, the middleware and
database systems that support them, and the
infrastructure they run across, including rout-
ers, switches, load balancers, firewalls, WAN
optimizers, and even the UPS and environmen-
tal systems they depend on.
2. Don't Make People Go To Multiple Places To Gather
Information
3NETREO - How to keep your IT Management from becoming shelfware

Many network management platforms fall
victim to the 'show them everything' fallacy.
It's common knowledge that putting an over-
whelming amount of information in front of
someone causes them to miss important
data-points, because they stop actually inter-
preting the material and start merely skim-
ming it instead. Any software that defaults
users to a screen where there is a huge list of
log data to go through, or multiple sets of
unrelated information, is going to be harder to
use and slower to resolve problems than one
that presents a clean, uncluttered overview of
everything that's going on. Requiring an
engineer to navigate 10 levels of folders to find
out what's turning that light red is just making
their job more difficult and increasing the
chances that it will impact the users before it's
resolved.
In addition, a simple to use interface will tend
to perform better, especially on mobile plat-
forms or when trying to access the system
remotely. Often engineers are on-site trying to
help resolve a problem and forcing them to
download and install new software on a
remote computer that they're using is a pain.
Simpler and easier means reduced training
requirements, which means faster deployment
and less time bringing the team up to speed,
which translates into quicker problem solving.
It can also help to provide 'self-service' views
for non-technical stakeholders into the state of
the environment, allowing managers and
executives to see that "Yes, the website is OK"
without having to understand the individual
naming conventions for all the servers, data-
bases, load balancers, and firewalls that go into
making that work. That means they're not
interrupting engineers to ask about the status
of things.
3. Keep The Interface Simple And Fast
It's important to consider the ongoing service
burden that implementing a management
system is going to create. The less daily
care-and-feeding a system requires, the more
likely it is to stay relevant to the business and
to actually solve problems for you. In spite of
this, most management platforms seem to
operate on the assumption that you have
several full time employees dedicated to
constantly keeping the thing running and
configured.
Look for systems that don't require you to
manually configure every change, as that leads
to rapidly out-of-date configurations. Good
management platforms should integrate
automatic discovery, synchronize with other
platforms (like your virtualization system to
detect new virtual servers, for example), and
provide an easy way to import new devices
without relying on your staff to decode
obscure SNMP MIB files in order to add support
into the platform. Ideally, you want a system
where patching and updating the software is
automatic, so it doesn't add additional operat-
ing overhead for your staff.
4. Minimize Administration

Schedule a periodic configuration review for
your management platform, at least twice a
year (more often in very dynamic environ-
ments) to make sure that as you add new
applications and systems, and the environ-
ment changes, that it is still monitoring the
things you want in the way you need them to
be - and that it's not generating excessive
alerts in doing so.
Another important thing to look for is the
ability to automate reporting, and to quickly
generate new reports as needed. If the soft-
ware requires someone to learn a new script-
ing language, or for a network engineer to
learn SQL in order to generate reports, it's not
likely to be used much. Standardize on systems
that allow quick ad-hoc reporting to fix prob-
lems as they come up and to address the
concerns of management or executives as
needed, but that also then allow those reports
to be automatically sent out. Busy engineers
will often forget to go look at reports, some-
times waiting on alerts or even users to inform
them of problems that could easily have been
prevented by getting an email once a week of
'servers with lowest free drive space,' for exam-
ple. Time spent manually creating reports
every month is time that could be easily saved
and dedicated to more pressing tasks - allow-
ing a smaller staff to handle a larger environ-
ment.
Buying software is putting a capital investment
into something that may not fit your business,
or your needs, a year from now.. or ten years
from now. Sure, initially buying the software
might seem to let you front-load your costs,
but don't forget what happens when you buy a
1,000 device license and then downsize a few
locations or push applications out to the cloud
a year or two from now and only need half of
that. And you also usually have to budget for a
paid upgrade every year or two, plus the labor
costs required to implement it. If the software
you select is not appliance-based (but instead
runs on a general purpose operating system)
you also need to carefully calculate the full
price of the software you're buying - remember
most software has a lot of hidden deployment
costs in the form of OS licensing, hardware
requirements, database licenses (often from a
third-party vendor), backup software, and
anti-virus licensing. It will also add to the
operating overhead of your IT team, adding
additional systems to patch and maintain.
Often companies will discount your initial
purchase, but base the recurring maintenance
price on the non-discounted price, which can
be as high as 25% per year.
Subscription-based software insures that
you're not paying for any more licenses than
you need, and that you always have the most
current version. And the lower annual cost can
help you stay in-budget, without 'first year
spikes' in pricing, so it's easier to migrate your
legacy systems.
5. Don't Buy Software
5NETREO - How To Keep Your IT Management From Becoming Shelfware

Netreo has been helping customers successfully deploy, implement, and manage our OmniCenter
network management appliance in their environments for 15 years, and we've managed to retain
more than 90% of those customers. A large part of the reason why is that OmniCenter avoids
becoming shelfware by addressing each of these items in an elegant, affordable way.
OmniCenter includes extensive tools to mini-
mize false and redundant alerts, including the
automatic detection of dependencies and root
cause, alert suppression, planned maintenance
windows, and advanced rule-based incident
management. These tools work together to
make sure that when you get an alert, it's
already been verified as a real problem and
that it's the root cause you should be address-
ing. It can even provide advanced information
as part of the alert, like a list of the users who
are using the most bandwidth at that location,
or customized alerts to tell you if a redundant
device has prevented user impact.
You can configure the sensitivity of the alerts,
including tuning the number of times to
re-check a service and the intervals to check at,
as well as setting time frames for thresholds (so
that small spikes in the CPU of a SQL server
don't generate alerts, for example). OmniCen-
ter also lets you filter alerts by time-of-day and
day-of-week, so if you want to avoid getting
alarms during your recurring maintenance and
backup windows, that's easy to do. And Omni-
Center is one of the only monitoring systems
that is time-zone aware, so all alerts and
reports can be automatically adjusted based
on the location of the server or devices you're
monitoring, insuring that "Working hours"
means the right thing even for your systems in
New York, or Dubai, or Chennai.
OmniCenter Automatically Minimizes Alerts
OmniCenter includes many pre-configured
dashboard views into every aspect of your IT
environment, including applications, voice,
email, cloud services, and virtualization, and
you can customize your own views as well.
Users can be assigned to partitions to insure
they only see the devices they are responsible
for, further de-cluttering the interface, and
important problems are right where they
should be - right at the top and never more
than 1 click away from detailed information
about what's causing any alert. OmniCenter
can even show you potential problems before
they happen, with failure prediction and
long-term trend analysis.
Designed from the ground up for ease-of-use,
OmniCenter's user interface ensures that even
with minimal training users are able to quickly
identify problems and get detailed information
about what's going on from anywhere in the
extended enterprise. Even from mobile devic-
es or tablets. Different types of devices are
presented in a consistent format, making
multiple-platform management easy. And the
entire interface is web-based so there's never
any clients or applets to download or install.
OmniCenter is designed to let you spend less
time‘managing the management platform’
and more time focusing on your customers.
One Pane-of-glass That's Easy To Use And Understand

Near-zero Administration
By automating most repetitive maintenance
tasks, OmniCenter reduces the routine mainte-
nance tasks required to near-zero, freeing up
engineering time. Intuitive design and default
settings allow for minimal administration and
quick deployment. OmniCenter also simplifies
user and permissions administration by inte-
grating with Active Directory, LDAP, or SAML
authentication systems. OmniCenter can even
open a secure VPN connection from your
network to Netreo, allowing Netreo engineers
to provide remote configuration support,
upgrades, patches, and even new device
definition or custom coding with minimal time
requirements for your team.
Reports can be created through the web
interface in seconds, and then saved as favor-
ites to return to with a single click, or automat-
ed and scheduled to go out automatically.
Hundreds of built-in reports and simple-to-de-
fine custom reports can be linked together into
report templates to address even complex
reporting needs, all without scripting or writ-
ing SQL, and without the need for any addi-
tional reporting software or client programs.
For the ultimate in easy administration, Netreo
provides managed services for OmniCenter,
allowing Netreo engineers to provide 'cloud
level' support for the platform, even hosted at
your site. This guarantees the management
system will stay in perfect sync with your
environment, including periodic configuration
reviews, alert tuning, and incident analysis.
Managed services free your staff to concen-
trate on improving your core business service
levels, without worrying about their manage-
ment platform.
Using our appliance-based architecture,
rollouts are a snap. You’ve got your choice of a
VMWare-based Virtual Appliance (VA), one of
our robust and redundant hardware applianc-
es, or a cloud-based appliance. All of these
options include all of the inherent advantages
of an appliance-based architecture: there is no
general-purpose underlying OS to administer,
no databases to license or upgrade, no anti-vi-
rus licenses to deal with, no extra hardware to
purchase or spec out. There are also no clients,
agents, or probes. Regardless of appliance
format, scaling is easy and seamless whether
you need to monitor 5 devices or 50,000.
A simple, self-contained solution with one low
monthly or annual cost that covers the full
OmniCenter license and includes free revision
updates. There are no terms, commitment
periods or contracts. Devices licensing can
change month-to-month with no penalties -
you pay only for what you need, so you're
never stuck with licenses you paid for but
aren't using.
OmniCenter is used by hundreds of leading
universities, government agencies, and public
and private enterprises of all sizes to monitor
and manage complex, mixed-vendor networks
with stringent uptime and performance
requirements. For more information or to start
a free trial today, please contact Netreo at
866-Netreo1 or online at www.netreo.com.
Hassle-free Purchasing And Deployment
7Enterprise-class Monitoring Made Easy

Netreo whitepaper 5 ways to avoid it management becoming shelfware

More Related Content

What's hot (19)

Viewers also liked (12)

Similar to Netreo whitepaper 5 ways to avoid it management becoming shelfware (20)

Recently uploaded (20)

Netreo whitepaper 5 ways to avoid it management becoming shelfware