The proliferation of log files, e-mail, CAD/CAM, ERP, and the shear size of computer application packages these days tends to consume a tremendous amount of disk space. Add to that the fact that new government regulations require certain businesses to store data for longer periods of time than ever before, and it's easy to see that without proper disk management policies, an organization can find its storage requirements growing out of hand at an alarming rate.
As the size of data volumes expand, the amount of application processing and response time grows right along with it. That's because applications that have to read and write data perform less efficiently when disk space is low. These inefficiencies cause performance backlog and slow down the entire network. Low disk space also impacts the ability of your server to grow the paging file which impacts virtual memory management. Of course, once conditions begin to deteriorate to this level, the "slow network" complaints start pouring in from the users.
Running low on disk space is much more than an inconvenience. When servers run out of disks space, transactions fail, web sites go down, and the repercussions are felt throughout the entire enterprise.
Network managers should use this 5 part strategy to stay ahead of disk problems:
Make sure that disk and file systems are configured for best performance.
Use available disk quota management systems.
Have a written data storage policy that defines what type of data may be stored on the network drives and how long it may be kept there, as well as other basic data management requirements like these:
Require users to save all downloaded Web content to their local drive.
Require users to clean their home directories at least once per month.
Require users to store e-mail attachments to their local drive.
Use disk space monitoring software to warn of a problem before it occurs.
Take corrective action to alleviate disk space problems immediately upon notification.
While most operating systems do have basic disk space usage reporting capabilities, they lack a full-featured, agentless low diskspace monitoring, reporting, and notification system that will watch for disk problems 24/7. This is especially important if your NOC is not staffed around the clock. There are application that have been known to consume all available disk space when they hang up or crash, and then there are virus, Trojan horse, and network attacks that can take available disk space down to 0 bytes. If a live manager isn't there at all times, you've got a potentially serious problem on your hands.
If you don't use disk space monitoring software, or if the software that you do use is not capable of tracking you down by e-mail, cell phone, pager, and instant messenger to tell you that you have a low diskspace problem, then it's time to get some.
Your best bet is to cover all of your bases and go with a complete network monitoring package that will also watch your web servers by checking all of the various protocols and services that it's running, and knows how to monitor and respond to SNMP Traps.
The better packages are rules-driven and even allow you to designate certain remedial actions for the network monitoring software to automatically take whenever an alarm occurs.
Installing network monitoring software endures that you have a safe and reliable way of knowing when disk resources are low, as well as a 24 hour digital sentry, standing guard over your entire enterprise, who will instantly notify you if anything is amiss.