Comment to our previous blog:
Ontonix is launching a huge change and I love changes. Things that do not change scare me, because they look too much like death. However, for years we have been looking for resilience, we have been defining resilience, we have been measuring resilience, we have been preaching resilience.
Are we now disavowing years of work and beliefs? Are we betraying resilience in favour of anomaly detection? I do not think so.
Let us summarize:
Resilience is heavy
True. The measures adopted to mitigate the vulnerabilities of a system by increasing its resilence add weight to it and “plaster” it, resulting in a limitation of its agility, thus turning the apparent strength of its sturdiness into an additional weakness.
Resilience is expensive
True. The features added to improve the resilience obviously increase the cost, at least during the design and manufacturing phases.
Resilience is static
True. Resilient systems are resilient against change too, while the world outside changes quickly, continuously offering new challenges, so a safe and robust system today will not necessarily be safe tomorrow.
Resilience is potentially unsafe
True. Despite its immunity to current threats at a certain point in time, a resilient system may prove to be highly unsafe one moment later, due to the inherent resistance to change, which makes its adaptation to a mutated scenario harder and, above all, slower.
However, resilience is a property belonging to systems. Anomaly detection concerns the flows, therefore lies in the field of processes.
Consequently, shifting the emphasis from systems to the processes is not a U-turn, but a progress towards a more efficient application of the benefits provided by the same strict and scientifically based measurement method, which is Quantitative Complexity Management.
It is a paradigm shift, which consists in moving from the domain of systems to the domain of processes, therefore focussing on anomaly detection, which is agile and flexible, instead of resilience, which is heavy and rigid.
According to Darwin’s evolutionary law, the most successful living species are not the strongest ones, but those with the greatest adaptability. Resilience is still the key evaluation criterion for calculating the reaction capability of a system now or in the near future, especially against known threats. However, we realized that the most effective option to plan and keep updated the protection strategy of a system is to base it on fast anomaly detection.
Posted by MdL