Power And Value Of Defragmentation
Some of the leading analysts have stated that NT/2000 sites
worldwide are wasting their budgets each year on unnecessary
hardware upgrades. In a recent report by International Data
Corporation, many of these upgrades and purchases are an attempt
to solve a performance degradation problem that is actually
caused by something unsuspected: -- disk fragmentation.
Steve Widen, Director of IDC's Client Infrastructure Storage
Software Research said: "Because fragmentation exerts a severe
toll on system performance, it's quite likely that many organi-
zations have initiated hardware upgrades unnecessarily," He
continued with: "IDC estimates that network defragmenters can
save corporations $6 billion per year by preventing unnecessary
hardware upgrades."
Instead of attempting to remedy the situation by buying new
machines, a new IDC report, entitled "Disk Defragmentation for
Windows NT/2000: Hidden Gold for the Enterprise," recommends
regular defragmentation of every server and workstation across
an enterprise. You can download this from our Diskeeper page
below.
Keep Systems At Peak Performance
IDC came out with the following very interesting calculation:
"Enterprise-wide deployment of defragmentation software reduces
desktop TCO by around $350 per year per computer". These are
definitely figures that need to be communicated to your MIS
management, and remember to mention the fact that regularly
scheduled defrag runs maintain your systems at peak performance
and prevent you from having to rebuild systems that are getting
slower and slower over time.
Bob Lofasco, Lucent Technologies' Manager of IT said: "We
experience at least a 15 to 20 percent performance gain on both
servers and workstations by defragmenting our systems daily."
Normally, Sunbelt sells all products all over the planet, but
in this case, we have TWO different solutions. For the USA and
Canada, we provide SuperBoost-NT which includes the Norton
SpeedDisk defragger, a tuner and a data cache tool. For Europe,
we provide Diskeeper for NT and W2K. Check out the specs:
Europe:
http://www.sunbelt-software.com/product.cfm?id=455
USA/CA:
http://www.sunbelt-software.com/product.cfm?id=900
Data Deluge Drives Storage Management Innovation
Noticed recently that a decent size hard drive does not last
that long anymore? The need to store more data is exploding
with the 'Net and bloatware that takes up more than one CD
just in executables and dll's.
Storage is rapidly evolving as a specialized function and often
needs more than just one way to get managed well. For instance
Cisco Systems and Brocade announced that they are going to be
able to connect their Storage Area Networks (SAN's) to each
other over IP-based Metropolitan Area Networks (MAN's) and over
Wide Area Networks (WAN's).
That is why one of our vendors WQuinn has been working hard
to continue building their StorageCeNTral product. You will
need more powerful software tools to manage all this stored
data. StorageCeNTral allows you to manage all your data through
a single, powerful storage management console, leveraging the
company's roots in the VMS storage realm (Large mini-computers)
into the NT/2000 systems world that MS is not making easier.
Useful performance data removed in Windows 2000
Under Windows NT up to version 4.0 in performance monitor lived
the very useful performance object "Logical Disk". It allowed
you as the admin, for instance, at a glance to check the
relative and absolute free space of all disks on all Windows
NT machines in his network, without the need to map every disk of
every machine. Under Windows 2000 this object "Logical Disk" has
been removed. The admin has a problem when he urgently has to
find out all disks which run out of space in his network.
StorageCeNTral allows you to keep disk management chores down
to a minimum, assign quotas in real time, get fast and instant
space updates from your SAN, WAN or SCSI server disks and much,
much more than I can mention here. Some of the really cool things
are the customizable Report Writer and Capacity Planner. This is
a tool you really should check out as it is a MAJOR timesaver.
Last but not least, a smart way to present the disk quotas concept
to your users is the following positioning: "We reserve your space
for you". In other words, we make sure that nobody else is a
space hog and threatens _your_ disk space. Looking it from that
perspective might even make them ASK for disk quotas.
http://www.sunbelt-software.com/product.cfm?id=423
Business Continuity Planning
DEFINITION: Companies use this to make sure they are able to
continue to operate in the event of a technical failure or
other emergency.
These days, if the power goes out, a virus hits or a system
crashes (we just had one of the nodes in our web-cluster do that
as it ran completely out of disk space) the results can be
catastrophic. Crucial data is lost, profits are stalled and if
it is not solved quickly, business grinds to a screetching halt.
Unless you were prepared and had a business continuity plan in
place for such an emergency.
Lucky for us we had a cluster and used Double-Take, so the other
node kept things going. But it illustrates clearly that we are
living in a 24- by-7 world.
The old 9-to-5 is no longer. The Web makes it 'round the clock
and _many_ NT/2000 servers are intranet or web machines. That
means business continuity becomes far more important. If for
instance Nasdaq would fail, it would disrupt the financial
markets and the world economy to some extent.
Business continuity plans make sure there are always backups
in place that can kick in and take over the load. This takes
many forms. Making sure that your UPS is really uninterruptible,
and that there is a backup for that UPS is just an example.
But safeguarding your data is even more important in case a
hurricane, flood, or other natural disaster wreaks havoc on
a particular area. You need to have your data off-site in
such a case. Your business continuity is too often an after-
thought. It's always better to prevent the disaster in the
first place.
You basically cannot afford to wait until you are faced with
a critical event to address the issue. You need to do it
ahead of time. You need to accelerate and perhaps formalize
the effort more. Grab your colleagues by their collar into
a room where you can sit for 30 minutes and do this: "We got
hit by a flood from the fire at the business next door. All
servers are fried. How can we keep on going?"
If you have NO IMMEDIATE ANSWER, it is HIGH TIME you check
out Double-Take, and the Microsoft Web Seminar that explains
how to do Disaster Recovery featuring Double-Take. All here:
http://www.sunbelt-software.com/product.cfm?id=111
(This was inspired by an article in Computerworld June 26, 00.
Grateful acknowledgements.)
This issue of W2Knews™ contains: