A better title would be “Life-inspired applications”. But this term has already been used, so I used this one, although it may wrongly suggest a change in hardware technology.
Thinking about the future of IT we cannot forget the service continuity.
In the current IT systems – especially the big ones – there is always someone responsible for service continuity – whether it is the organization’s department, or outsourced service based on SLA.
When the organization interoperates electronically with its environment (partners) it extends the scope of service assurance with external agreements defining mutually acceptable SLA conditions.
This approach of ensuring business continuity based on internal and external agreements is currently effective because:
- The majority of the e-business flows runs within the organization and not in-between.
- The number of electronic interoperations is countable and manageable in the organization
- The organization is aware of the identity of the external service provider
We can imagine in the next decades, none of the criteria above will be valid in future e-business-service-mesh.
We can imagine that even the critical business processes will be run within an IT system, which will rely on external services, with several providers available (look at DNS service providers to understand).
There will be no closed list of service providers to set agreements, and there will be no willingness nor acceptance for a guaranty of service availability.
I can’t predict whether the SLA’s will disappear or change.
What I know: having unstable environment, and the lack of centralized SLA management, the applications will try to become self-manageable which means, they will behave like organic systems in terms of:
- Auto-diagnosis
- Auto-repair
- Auto-scale (this is partially happening with autoscalers)
- Auto-reconfiguration
- Auto-adaptation (partially happening with circuit breakers)
I ‘am sure there many other autos, I haven’t mentioned here.
So the challenges for IT engineers have changed:
- At first they tried to write a program that runs without errors
- Second, they tried to write a program that makes what it should
- Third, they tried to write a program that is able to be serviced and managed
- Next they must learn to write a program that will be able to auto-service and auto-manage
But this is a challenge for the new generation of engineers (or GPT successors 😊 ).