Automation May Be Your Salvation

When I interview IT professionals for articles I'm working on, I often try to gauge just how pressure-filled their work environments are. I'm particularly interested in hearing the responses from folks at small to midsized organizations, where budgets can be tight and IT staffing limited. Of course, asking them about their workdays requires that I actually get them on the phone - no easy feat sometimes. When the person's primary job responsibilities are conducted in and around a data center, it's not unusual for me to phone in at the scheduled time only to get the voicemail system. Several hours (or days!) later, when the person returns my call or actually picks up the phone, the explanation is familiar - something along the lines of "Sorry I missed your call. One of our UNIX servers went down, and we had a bunch of people who couldn't access the database. It took us the rest of the day, but everything seems to be OK now."

And, as I listen to these stories of crisis management, I often ask, "So, just how many of you are there?" The answer to this question is also familiar. A typical IT staff for a small to midsized enterprise is 4 to 5 people - sometimes as few as 2, rarely more than 10. I recently spoke with the IT director for a midsized university. His team manages a multiplatform server farm of nearly 100 servers, running all of the college's business and education-related applications and databases, including file servers for the home directories of the institution's more than 10,000 students. The small SAN (storage area network) they recently built already houses nearly 2 TB of data and will, within a year or so, hold 4 TB. When I asked how many IT people support this rapidly scaling infrastructure, he answered, "Well, there's me ... and the two people who work for me."

Set Storage Policies And Then Sit Down
Can companies continue to manage their growing IT environments with relatively small numbers of human resources? Yes, but only if they give their IT staffs tools designed to reduce the time spent on direct server intervention down in the data center. Given the rapid proliferation of applications and data, even in small to midsized organizations, small IT staffs can't monitor all data center equipment at once - not without help, anyway. Of course, it can be tough to convince the higher-ups (i.e. the holders of the IT purse strings) to cough up cash. After all, if money were no object, the IT staff would have the dozen or so people it could really use. But, short of spending the money to double or triple the size of the IT team, the organization should make an investment in automation - software that takes over many of the day-to-day systems management duties. What overburdened IT staffers really need is the ability to let the machines watch the machines.

Perhaps the best place to start is in storage management - that's where explosive growth in data most clearly threatens uptime. Overload your servers and storage devices, and watch your applications come crashing down. Various software vendors now offer increasingly sophisticated tools for monitoring and managing storage capacity and usage. SRM (storage resource management) tools, in particular, allow companies to set policies on what types of information can be stored, who can store it, how much, how often, and on which devices. The tools also send out automated alerts when storage capacities are reaching their limits or when poorly allocated storage resources are degrading application performance.

Most importantly, SRM software can automatically reallocate storage resources as application and user needs fluctuate. Combined, these features can greatly reduce the frenzied moments when IT workers find themselves hopping from machine to machine, trying to keep the lights on in the data center.