21.3.4. Proportional Allocation Scheme

Now we look at schemes for the allocation of frames. Till now we have been looking at schemes where I have the entire set of frames in memory and any page can be replaced by any process. This may hit the performance of processes; to avoid that, each process needs a minimum number of frames. For example, to execute an instruction, it will need one page which contains that instruction. It will also need pages containing data it may need to read or write. Therefore, each process at any given point in time will require a minimum number of frames to ensure a minimum level of performance.

In this introduction, we learn about frame allocation schemes. A frame is a fixed-size block of physical memory that stores page contents. To allow processes to run efficiently, a certain number of frames must be allocated to each process because they need pages for instructions and data. If processes don't have enough frames, they are more likely to experience 'page faults,' which slow down execution as the system needs to retrieve pages from slower secondary storage.

Think of a library where each student (process) needs a certain number of books (pages) to study. If a student only has access to a single book, they may frequently have to switch between bookshelves (secondary memory) to get the information they need, making their study time inefficient. Allocating enough books allows them to study without interruption.

Now, allocation schemes are typically of two types: fixed allocation and priority-based allocation. In a fixed allocation scheme, all processes get the same number of frames, whereas, in proportional allocation, frames are distributed based on process size. This approach aims to allocate frames in relation to the needs of each process rather than treating them equally.

Fixed allocation divides the total number of frames equally among all processes regardless of their actual needs. However, this isn't always fair since different processes might require different amounts of memory depending on their size and complexity. Proportional allocation addresses this by analyzing the size of each process. Larger processes get more frames while smaller ones receive fewer, ensuring that memory resources are allocated based on demand rather than a simple one-size-fits-all approach.

Consider a classroom where all students (processes) are given the same number of tools (frames) to work on their projects. If one student has a larger project that requires more tools than others, they will struggle to complete it efficiently. If instead, the teacher gives tools based on the size of the projects, each student can work more effectively at their own pace.

In proportional allocation, the number of frames each process receives is calculated based on the size of that process relative to the total size of all processes. For instance, if there are 100 total frames and one process requires 10 pages while another needs 127 pages, we can determine how many frames to allocate to each based on their sizes compared to the total size of processes.

The calculation for proportional allocation is as follows: if we denote the total frames as m, the size of process P as s, and the total size of processes as S, then the number of frames allocated to a process can be calculated using (s/S) * m. This ensures that processes are allocated a number of frames congruent to their needs, preventing larger processes from being starved of necessary resources.

Imagine a pizza where each slice represents a frame. If one person has a bigger appetite (a larger process), they should receive more slices than someone who is not as hungry (a smaller process). This way, everyone is satisfied without anyone hogging all the pizza.

In a priority-based allocation scheme, we can cater to high-priority processes by allocating more frames to them. This ensures they have sufficient resources to operate effectively while lower priority processes might receive fewer frames. Additionally, a proportional scheme can be implemented taking into account both the process size and priority.

By implementing priority-based allocation, the system can ensure that critical processes receive the resources they need to function optimally. This is particularly important in environments where certain computations must be completed quickly or without delay. Using a mixed approach of size and priority allows the system to be more flexible and responsive to changing demands.

Think of a hospital where certain patients require immediate attention based on the severity of their condition (priority). While all patients (processes) need care, a patient in critical condition may require more resources (frames) so that doctors can quickly address their needs, while patients with less urgent issues can wait longer.

When frames are allocated to a process, allocation can be 'local' or 'global.' In local allocation, a process can only use frames specifically allocated to it for replacing pages. In global allocation, any process can use frames from the overall pool of available frames allowing for more flexibility and utilization of memory.

Local allocation restricts the scope of frame use to enhance the predictability of performance for each process, but may lead to inefficiencies since processes cannot access additional frames if facing issues like page faults. On the other hand, global allocation allows for a more dynamic adjustment of frames, enabling the system to adapt to varying demands, but could lead to higher contention and low priority processes being starved for resources.

In a restaurant scenario, local allocation resembles a table with a limited number of diners, only serving the guests seated there. But global allocation is like a buffet where all diners (processes) can access any available food (frames), leading to higher customer satisfaction but potential crowding.

--