Enterprise architecture driven by a model (Category Architecture)
I am interested in architecture in all its manifestations, so sometimes I look at the efforts of corporate architects. At the recent Sber conference, it was speech Oleg Zakharchuk on the theme of architecture controlled by the model. Oleg himself finished MIPT in the distant 1987 During the year, and now he manages ASIS Soft, is an expert in the field of enterprise architecture and a corresponding member of RAEN. He is known for his work in the creation of metamodel.
If you briefly summarize the ideas of the speech, it tells about
- The concept of holarchyThis will replace the traditional hierarchical structure. n
- The four-level modelactions, tasks, working groups, organizations
- Unified ontology of activity Creating digital models of enterprises based on this model
- Digital agentsthat must be somewhere huddled at the level of action and individual services to automate them (Apparently, this is a reference to AI agents.)
In general, if we try to analyze the problem areas of the approach proposed in the report, then 1. Conceptual complexityDespite claims of simplification, a four-tiered holarchy with multiple links of “commitments” may prove more difficult to understand and implement than existing approaches. 2. Scalability of the solution at the company level. Although the author argues that the approach allows describing systems of different scales, practical experience shows that unified metamodels often face problems in scaling to the level of large enterprises. 3. Integration with existing solutions. Most businesses already have an advanced IT infrastructure built on a variety of standards and methodologies. It is not clear how this model can be integrated into an existing company. 4. Support and development of the model. Unlike open standards such as TOGAF or ArchiMate, which are developed by international consortiums, Zakharchuk’s model is a proprietary solution of one company.
Next, I will share my opinion on the approach of the author, who reminds me of the sermons about the author. luminousIn front of an old man's corps. The author takes the approach from the past:
- Concept. holarchy was formulated in 1960The Philosopher Arthur Koestler
- The idea of one. ontology Companies are actively developing semantic web two-thousand
- And so on. The author rebranded and began to sell them as a working solution ... and everything seems good, but it does not work and this is a problem not only with the model of Oleg Zakharchuk, but with the entire approach to the architecture of the enterprise, where you can highlight the following problems:
1. Lack of connection with reality The general trend of enterprise architecture is to create beautiful theoretical models that don’t work well in practice. Priority is given to architectural purity rather than viable and effective real-world solutions. n 2. The problem of modeling for the sake of modeling Teams create models of something, but do not have a clear understanding of the purpose for which they are modeling. 3. Corporate architecture frameworks are useless. Frameworks like Zachman, TOGAF or FEAF are nothing more than typical management fads, aggressively promoted by consulting companies and gurus. They are useless at best and harmful at worst. (more whitepaper) 4. Pseudoscience of corporate architecture. Systems thinking, as a tool for enterprise architecture, is often praised in publications related to enterprise architecture, but in practice fails to provide any viable proposals. Popper's falsification (more whitepaper)
P.S. When I hear the word ontology in the context of architecture, I realize that things will only get worse.
#Architecture #Software #Engineering #Management #DistributedSystems