Given the complexity of managing IT environments these days, it’s now only a matter of time before machine learning is routinely applied to manage IT operations. One of the first companies to provide such a capability is SIOS Technology, which today announced the general availability of SIOS iQ software for VMware environments that applies analytics based on machine learning algorithms to both IT infrastructure and application software.
Available in both standard and free editions, SIOS Technology COO Jerry Melnick says the machine learning software first automatically discovers what should be defined as normal within any IT environment, and then over time learns what deviations from normal will result in a particular performance threshold being broken or potential vulnerability being created.
Melnick says SIOS Technology decided to focus initially on the VMware environment because of the size of the installed base, but the technology will soon be more broadly applied. At its core is an implementation of a Postgres database running machine learning software that IT organizations download onto a VMware virtual machine. Via a SIOS PERC Dashboard, SIOS iQ then recommends the best solution to any particular issue it discovers.
Priced at $150 per host annually, SIOS iQ also provides analytics for Microsoft SQL Server running on VMware, host-based caching that helps optimize storage I/O performance, and the ability to identify and eliminate idle VMs and snapshot sprawl.
The challenge with most existing approaches to managing IT operations is that most of the current software used in these environments generates far too many false positives. The end result is that the IT staff winds up either chasing its tail or becomes so inured of the alerts that they stop paying any attention to them at all until a crisis ensues.
Naturally, not every member of the IT operations team is going to be enthusiastic about the level of automation that will soon be enabled by machine learning software in even the most complex IT environments. But like it or not, machine learning in the enterprise will soon be here to stay.