Managing Disk Storage Can Turn Windows NT/2000 Servers Into Top Performers

Any employee can compromise server performance and application availability by storing large files on servers shared by other employees. Directory disk quota thresholds can minimize downtime and enhance performance.

By Elizabeth Ferrarini | Posted Dec 21, 2000
Page 1 of 2
Print ArticleEmail Article
  • Share on Facebook
  • Share on Twitter
  • Share on LinkedIn

During a recent hurricane warning, the file servers at the County of Miami's office in Dade, Fla., began crashing before the first drop of rain hit the beaches. It wasn't the hurricane's high winds, or the related flooding and power outages that caused these failures: It was the county's employees.

Fearing the worst for their desktops after being hit three years in a row by hurricanes, employees developed the habit of copying their desktop computers' hard drives onto the servers as soon as news about a new hurricane hit the airwaves. Dozens of employees, with desktops averaging four to eight GB, started to copy their desktop data to the servers simultaneously. The servers couldn't take the pounding.

Not every IT organization has hurricane warnings to worry about. On the other hand, every Windows NT/2000 server systems administrator needs to ensure server availability and performance. Specifically, they need to be concerned about uncontrolled server disk capacity consumption by employees.

Strategic Research Corp., a storage research firm in Santa Barbara, Calif., estimates that disk drive failures in distributed Windows NT networks cause about 55% of all server downtime, as illustrated in Figure 1. This percentage includes failures by servers that have exceeded their capacity. Strategic Research Corp. also estimates that disk space consumption in some organizations has reached rates from 70% to 100% annually. Overloaded capacity can make a server's performance suffer.

Overloaded Capacity Can Lead to Failure

Figure 1
Figure 1: Disk drive failures cause the majority of server downtime

Short of a complete server crash, any employee can seriously compromise server performance and application availability by storing large files on servers shared by other employees. Some systems administrators deal with this problem by manually babysitting a server's capacity level or using load balancing to control the amount of data on a server. A better solution is setting multiple, real-time directory disk quota thresholds on the amount of space assigned to each employee and to each department. Employees should receive automatic alert notifications as they get close to their quota thresholds. This procedure allows systems administrators to allocate storage evenly as a shared resource, and to make sure the types of documents saved conform to the IT department's policy for what belongs on the server.

Use Soft Quotas to Monitor Server Disk Usage Transparently

Directory quotas enable systems administrators to maintain server performance very easily, and to avoid system crashes due to exceeded capacity levels. Soft quotas allow employees who've exceeded their quota threshold to continue to save files on the server. Systems administrators can automatically execute command lines, set alerts to be triggered, run reports, or send the activity to the event log.

Soft quotas provide a good way to monitor the status of any server's disk usage in relation to the server's disk capacity. Alerts can provide plenty of warning when a server is about to reach its allowable capacity. Soft quotas are useful for monitoring server disk usage at the 50% to 85% percent range.

On the other hand, soft quotas can work only if the IT department educates and empowers employees in grooming their disk space following a quota threshold alert. Otherwise, quotas will have limited effect on employee's storage usage.

Comment and Contribute
(Maximum characters: 1200). You have
characters left.
Get the Latest Scoop with Enterprise Networking Planet Newsletter