Helpful Tips
Get to the information you want quicker by selecting a category name or a popular tag.

Any service, whether consumed or delivered must be monitored. As mentioned by some previous posts, the monitoring measures against the established norms as a judge of whether the service level agreements are being met. Outsourced or in-sourced does not matter. An organization has a responsibility to internal and external customers that need to be delivered upon. Monitoring is a proactive way to understand the health of the ‘services’ being delivered which, in turn, allows for adjustment, remediation, and risk management. 

Each organization will measure against different metrics based upon the culture, vision, and goals of that organization. No software, hardware, or technology can blindly and out of the box deliver upon an organization’s metrics realistically; unless you truly feel your organization has no compelling differentiation against the competition. Ironic that some feel that they are no different than others in the same industry because the fact that you are competing in the industry states the fact that you are unique in some form or fashion. Therefore, each organization has unique metrics upon which they need to measure success, failure, and sub-par delivery of ‘services’ offered to their internal and external customers.

As a start, I usually give the following metrics a detailed discussion within the culture of the organization I am dealing with:
* Performance – I/O requirements for response time, etc.
* Availability – local availability requirements for uptime, maintenance, etc.
* Reliability (MTBF) – mean time between failure
* Connectivity – connectivity requirements – protocol and medium dependencies
* Scalability – growth potential
* Information Criticality – accessibility criticality over life of information
* Recovery Priority – operational recovery requirements (who, where, why, etc.)
* Retention – defined retention categorical stratification – may be tied to compliance
* Compliance – regulatory concerns such as SOX, NYSE, NASD, SEC, FCC, FTC, OCC, GLB, VISA, Basel, etc.
* Data Integrity – security level for creation, access, retention, and deletion respective to transmission and at rest

Notice that I have distinctly stated ‘services’. Too many organizations attempt monitoring without establishing services, service level agreements, or even a vision of where they are going. For monitoring to be meaningful, these things must be established prior to the ‘monitoring’. Otherwise, there is no understanding of what is being monitored for. Measuring up-time, CPU utilization, I/O thresholds, and so forth is meaningless, even for capacity planning, without the construct for operation that services, service level agreements, and goals (targets to accomplish the vision) provide.

Leave a Reply

You must be logged in to post a comment.