The Need for Monitoring and Management

When mission-critical computing was confined to mainframes, operating systems typically had built-in facilities that monitored the system’s health, and application vendors could take advantage of these facilities to monitor their application. When a job failed, an operator would be alerted on a console and often given guidance on what to do next.

When distributed systems first became popular in the 1980s, they most commonly provided file and print services or e-mail services for a small group or department. If a system had a problem, an end user would discover it and call the help desk or a technician to investigate and correct it. Although such outages might indirectly impact the business, they were usually limited in scope and cost. Centralized monitoring tools, while potentially useful, did not have enough cost justification to spur wide-scale adoption.

Fast forward to today. While mainframes have not gone away, nearly all medium and large organizations use Unix and Windows servers to perform mission-critical functions. Furthermore, the e-commerce revolution means that employees are not the only ones affected by outages; customers and partners also depend on systems to be healthy. Outages can be extremely expensive in both immediate financial loss as well as eventual loss of confidence and market share.