The Automated Data Center - Still A Dream?
A recent article in CNNMoney talks of a recent study which found that while virtualization is increasing in the enterprise, automated management tools to improve efficiency and IT resource utilization are being implemented by relatively few. Taking a more ground-level view, this is a dangerous trend. Having to manually operate data center management activities and include long personnel training curves on tasks right in with the day-to-day expansion and putting out of fires is clearly an invitation to disaster.
An IT staff, as anyone who has been there knows, already has a very overfull plate. When routine tasks are added to the implementation and debugging of hardware and applications and the handling of routine emergencies, it means longer hours for—and lengthening response times from—data center personnel. And when a disaster such as a power outage or a sudden malware infection occurs in such circumstances, it can bring an enterprise to a dead stop.
Lack of automatic management also affects personnel efficiency. As a company grows, so does the IT staff. Each new hire, in general, knows less and receives less training than experienced personnel. That leads to inefficient - and incomplete - knowledge transfer, which means that tasks take longer and, again, disaster recovery time is doubled, tripled or worse. The more routine management tasks that are automatic, the shorter the training curve and the faster the response.
Just a few of the routine activities which should be fully automated are:
– Upgrades and patch distribution and installation.
The technology has long existed for updates and patches to be automatically downloaded from vendors. Taking this one step further, routine updates, and patches, after downloading from vendors, can and should also be distributed and installed automatically, with no required attention from IT personnel or users.
– Backups.
There should no longer be a need for any attention from datacenter personnel on incremental or even full backups, save to periodically check backup reporting to ensure that the right data is indeed being backed up.
– Defragmentation.
Scheduling defragmentation runs should now be a thing of the past. Defragmentation solutions which require no scheduling, which run invisibly, in the background and with no impact on system resources exist and should be used in all enterprises.
– Disaster Recovery.
While there is certainly no “magic bullet” solution to disaster recovery, many of the routine tasks of bringing systems back online can be fully automated to make recovery as short and as doable (by any personnel) as possible.
– User preferences and tendencies.
Although a bit more involved, solutions exist which will track user tendencies and preferences with databases and other applications. Efficiency can be increased all across an enterprise when these preferences and uses are automatically tracked and applied to updates, upgrades and even new applications.
– Resource Allocation.
Which resources are assigned to which applications and at what times should be fully automated and should be responsive to circumstances so that unneeded frantic phone calls from users do not interrupt other IT activities.
It well behooves IT personnel to begin a regular program of fully automating regular tasks, beginning with the ones which seem to occupy the most time.
Tags: automated, automatic management
Leave a Reply
You must be logged in to post a comment.