Storage Resource Monitoring (SRM) technology was conceptualized around networked systems that communicated with dedicated software installed on an individual host machine. This architecture eliminated “agents” and used “collector” technology to manage the communication between the host and the SRM application. As networks turned complex, debates around efficiency of agent-based and agentless software became common.
While agent based storage monitoring architectures require an agent to be installed on each host or element in the environment, agentless architecture only requires the installation of “collector” software on a central machine. Agents run like daemon or windows services and execute a set of privileged commands to generate pertinent system information that can be displayed on the system or transported to a host SRM application server.
The collector in the agentless architecture acts as the command centre and uses standard protocols such as SSH or WMI to collect and execute commands over the network connection. The agent software gets updated when the SRM application gets updated automatically if it is configured to do so. All elements in the agentless architecture stack that are interfaced by the collector are driven by a single set of commands and rules that are configured while setting up the collector software. Updates are to be done only on the “collector software”.
Newer technologies have gone a step further and divorced data parsing and interpretation logic from the collector and have stored it as rules definitions files (stored on the Web hosted database or on the local database accessed by a portal that has access to update the elements), so that these files can be updated without triggering off change control events.
Agentless software resource monitoring is simple and does not have a large footprint on the technology stack. Many of the storage management tools are getting integrated with system management tools so that IT Administrators can monitor the storage and manage the network from a single central console. It helps enterprise win the battle with growing volumes of data in that these tools reveal the how, where, when and why of data volumes.
The benefits of agentless SRM are many.
- The tools are modular and organizations can choose to implement specific tools for specific needs
- The tools are considered to be decision support systems
- They can help IT administrators be proactive instead of reactive
- These tools make it easier to plan for any technology changes that you may be considering such as thin provisioning
- They can help justify the expenditure and give you a return on investment within 12 months of deployment