Reliability in processes. Part 2

1 Terminology

IN “Reliability in processes. Part 1» [OpRes24-1] were defined (simplified): reliability, process and reliability in processes. Reliability is the ability to operate without failure (work without failures). Process reliability is both the ability to operate without failure (“the main process”) and to produce the required result (“the main result”). Quantitatively, this is the probability of failure-free operation (the probability of finding the process operational) and the probability of the required result at the output of the process.

For recoverable systems, and processes are mainly recoverable systems, use availability factor – the probability at an arbitrary point in time to find the system (in this case, a process) in a working state.

This applies to both the IT system (cluster of servers) and non-IT system (safe), as well as to the system of processes (operations) and components of the process, including its resources.

If in the classical theory of reliability (Reliability in technology [27.002]) usually considers internal destructive factors affecting a technical system such as failure \ equipment failure \ software, then “Reliability in processes and operations” (operational reliability) also considers unintentional personnel errors (operational risks) and internal fraud, external attacks on the processes of the company and clients companies (social engineering), natural disasters (man-made disasters). Ultimately, it doesn’t matter: the “process system” failed (did not complete the task) due to some kind of breakdown or due to its overload (from an unexpected “influx of clients” to a DDoS attack), so we add Reliability in Processes to the indicators “availability»:

Availability factor (Kg) is the probability of finding the system in working condition (nothing has fallen off), and
Availability factor – the probability of finding the system in a free (unloaded) state, i.e. ready to accept and process the load (i.e. the process\system is not loaded to its limit).

The system must be in an accessible state efficient.

In addition to readiness and availability, consider the effectiveness of the process. For a truly reliable process, it is important not only that it itself does not break or be broken (the security and fault tolerance of the process) and that its performance is sufficient. It is equally important that the result of the process (operation) is what was intended, i.e. result (product, product, service) of the required quality and on time. Quality is usually fixed in technical specifications (TS) for the product, i.e. in “Technical Specifications for the Process”. Thus, quantitative indicators of process reliability, in addition to assessing the ability to perform something (action), should assess the probability of obtaining what was really needed (within the framework of technical specifications).

Similar definition:

The reliability of a technological process is its ability to ensure the production of products in a given volume, while maintaining the established requirements for its quality over time.
Thus, the technological system must be efficient both in terms of quality and productivity.
Accuracy is the property of a technological process to ensure compliance … of the values of the product quality indicator with the specified …
Stability is the property of a technological process to maintain the quality indicators of manufactured products within specified limits for some time.

Another definition from GOST R 53736-2009: Electronic products. The procedure for creation and production. Basic provisions:

3.1.19 reliability of the technological process: The property of the technological process of product production to maintain over time, within established limits, the values of parameters characterizing its ability to be reproduced while maintaining tune, stability, accuracy in regulated production conditions in accordance with the requirements of regulatory, technical and (or) design technological documentation.

Reliability in processes (process reliability) includes:

a) not only the reliability and fault tolerance of equipment (reliability in technology) and software, but also

a1) fault tolerance against human errors (“fool proof”, re-entry, etc.);

a2) cyber resilience, disaster resistance, etc.;

b) downtime of the process (system), caused not only by failures, but also by overload of the process (system), i.e. availability – as the ability of a process to process (accept) the load, i.e. not be overloaded and have sufficient productivity;

c) process effectiveness – the ability to produce a suitable result (quality product) at the output;

d) automated and non-automated processes.

A significant component in “Process Reliability” is played by:

– temporary reservation, because usually processes are extended over time;

– hybrid redundancy, for example, when non-automated processes (manual processing) are backup to automated ones (automatic).

2 Key reliability indicators in “Process Reliability”

In Fig. Figure 2.1 shows the relationship between process readiness and availability. The states “Good” and “Efficient” according to [27.002].

If the process is operational (or even functional), but the load on the process is so great that there is not enough capacity and customers remain unserved, then the process will not produce a valid product (result).

Downtime types:

– downtime caused by the inoperability of the process (its systems);

– downtime caused by overload of the system/process – either the production capacity has decreased, for example, one of the system nodes has failed (server – server cluster node, second cash register in a store, etc.), or the load on the process has increased, for example, unplanned influx of customers, which requires time to connect reserve capacity to handle the increased load (open a second cash desk).

It should be taken into account that in this case the term “load” (payload) does not coincide with the concept of “Load” from GOST R 27.301-2011 (destructive influencing factors):

A.1.10.1 Load resistance analysis is the determination of the ability of a component or product to withstand electrical, mechanical or other stresses and environmental influences that could cause its failure.

Although the exceeded payload also tries to “break” the process, i.e. This is also a kind of “negative” nature of the impact on the process (there are a lot of clients – of course, this does not qualify as “negative”) from the point of view of the reliability of the process, because leads to overload and denial of service, but still the nature of this load is different.