Data Protection Management:
Todaythe common approach to collecting information about computing activity across the enterprise is to use software agents installed on individual clients and servers. However, when this approach is applied to data protection management in the enterprise environment, it has proven both ineffective and inefficient. This is because agent-based technology cannot meet the following five requirements that define the viability of a data protection management solution for the enterprise: ease of installation; compatibility with existing products; change management; network impact; and system performance. Considered collectively, these five factors are daunting and require a well thought-out approach. But, agentless technology, although challenging to design, is the only approach that meets large enterprise requirements and enables data protection management success.
Data Protection Management Approach:
Software agents are applications that run in the background on individual clients and servers. An agent's function is to collect information in the environment. This may include configuration data, activity logs or real time snapshots of computing activity. Examples range from Simple Network Management Protocol (SNMP) agents, frequently used in networked environments, to everyday processes running in the background of a home PC.
In data protection management, virtually every reporting application uses an agent-based architecture to mine performance data and deliver it to a central collection place. Generally, these agent-based solutions have been well tested in controlled settings, and when they are installed properly on a dedicated backup server, they work according to specifications. However, when agent-based solutions are implemented in open systems across large enterprises, the complex nature of the environment often creates performance challenges beyond vendor control. Competition for network resources, an amalgam of software and hardware, and collisions with core application processes undermine agent-based solutions. Reliable data protection management for the enterprise demands a different approach.
Agentless Technology: Longer-term Investment, Greater Rewards
Agentless technology is quickly gaining traction in IT, specifically in patch management and application monitoring. However, it is still rarely used to support data protection management. Both practical and economic factors are behind this, as agentless solutions are time-consuming to develop. Most software companies would rather avoid the upfront investment and long development window required to deliver a feature-rich, robust agentless application.
With an agentless approach, the application resides on a central server and collects data by accessing monitored computers through standard file-sharing methods. The application requires file-system access, and it must automate a complex two-way dialog when authenticating each machine.
In spite of the challenges inherent in an agentless approach, in data protection management it is the only way to handle the millions of transactions generated by hundreds of backup servers in a large enterprise.
The Data Protection Management Dilemma:
Working with multinational corporations, in some of the largest data protection environments in the world, Bocada has gained firsthand knowledge about the limitations of software agents. For example, some enterprises have attempted to address data protection management challenging by implementing self-built solutions. While these self-built solutions may have successfully achieved basic capabilities, most of systems continue to be limited to a basic set of operations and partial-scale deployments.
Agentless technology, on the other hand, is an infinitely scalable approach to enterprise-wide data collection and is essential to successful data protection management in a large enterprise.
