What if you knew which business processes would be affected by a storage outage

The best representatives of the present generation of system analysis tools identify problems and show us which hardware is impacted. What we really need however, are tools that go one step beyond, which is to say tools that not only show what storage and other assets the problem affects, but also which business processes are impacted. In other words, if a particular array goes offline, we would know ahead of time which line of business is going to take the hit.

Software offering this sort of information certainly seems like a logical extension of existing analysis software. Sitting on top of and interoperating with the root cause analysis software, it would take the results of the analyses and apply them to a database that understands the business process dependencies associated with each IT process. Similar databases exist now – the topic of Configuration Management Database (CMDB) has a lot of visibility these days – but existing products focus on enabling effective delivery and correct management of IT services. No CMDB looks beyond IT process to provide any understanding of what goes on in processes that are outside the IT center.

When things were simpler and storage was directly attached to machines, and when each machine was “owned” by a particular department, none of this would have been necessary. If the accounting department’s machine went down, they were the only ones who need be concerned. But the shift to networked storage in at least one sense has become a double-edged sword: we have exchanged improved storage manageability and performance at the expense of distancing ourselves from the actual business processes that the storage supports. The result is that when a system goes down, the ripple effect can spread in all directions, and we are essentially blind to the effect.

A product that tracked business flow in terms of IT process would have some interesting implications. If you had one, not only would you be alert to the results of an unplanned outage, you would be able to use the data to understand the consequences of planned outages as well – and by some estimates, except in the flakiest IT shops planned outages represent the majority of system downtime.

For a while now some vendors have been producing modeling programs that show us which assets will be impacted when we change storage topologies, but these – as far as I am aware – lack any linkage to the business processes. All of this might more readily fall within the category of Business Service Management rather than storage, but these days BSM, network management and storage management are all on a crash course to convergence. Of the software that I have seen, BMC and Onaro seem to be making the most progress in this area, correlating changes with service level impact.

Right now service levels as reported in these programs are only associated with IT services; when the vendors take the next step service levels will link with business process. When that happens, and when the analytics take place in real time, you will have a better handle on what happens when a resource goes down unexpectedly, and will be able to plan preemptive maintenance in a much more non-disruptive manner.