Cloud native apps are being made more robust. This is being done using distributed systems and fault tolerance enabling them to sustain the failure of any single component. These apps are scalable, due to their segmented nature, enabling greater versatility of use.
This raises the question, why can’t IT management systems be Built this way?
It’s been argued that management systems require special care. They must be installed and maintained in a way that does not drive costs through the roof. There is a noticeable lack of innovation in these systems. This is probably due to the fact that they are not consumer driven products.
Why not design new IT management systems that are as flexible as the newest cloud-based apps? IT management systems that carry heavy workloads should be highly available by design. They should not force customers to configure HA and manually maintain the setup.
Management-systems should choose to Store Wisely
These systems generate quite a lot of data, not all of it is transactional in nature. Out of millions of stats that are generated, they can afford to drop a few. Customers spend a lot of time monitoring, tuning, and adding capacity to these databases. Managing them incurs massive costs in licensing and operational bottlenecks. Next-generation systems should be different.
Scale should be handled differently than it is now
IT management systems could be designed to be multi-site-aware from the ground up. This would include scenarios where the customer has a dual cloud strategy but wants a single management interface.
System components should scale in a Linear Fashion
The previous generation offered storage and networking in a single system, but one had to carry most of the load to support the other. This meant sitting on resources that can’t be spent and deteriorating assets.
Modular Design should be the norm
A customer may have a large number of clients making calls to the IT management system. The number of VMs under management may be low, but the volume of API calls can easily bog the system down.
The API service could be separate from the VM scheduling service, and both could scale independently without any user intervention. This could make a huge difference in operating costs. Without something like this, it could take weeks to sort out what the real issue and months to get your vendor to create a solution.
Most enterprises want their workloads and data to be on site and under their control. This doesn’t mean all the management components have to be on site. By splitting IT management control into two components, you can keep the main control plane on site. This would include compute, storage, and SDN layers. Meanwhile, the operations and consumption layer can run from the cloud. This is a huge advantage.
The core infrastructure piece does not change often. Customers request features that are mostly delivered from the consumption layer. Delivering this as an SaaS component makes these features available to customers in weeks or months, as opposed to years. An agile enterprise can now get management features delivered in an agile way as well.
The next-generation IT management system must be self-operating, freeing humans up to do what humans do best- delivering creative ideas.
Katrina a product specialist, solving issues for your computer server and power needs at RackSolutions.com