#Implementation of the Banker's Algorithm in Golang
The Bankers Algorithm is a resource allocation and deadlock prevention algorithm developed by Edsger Dijkstra that tests for safety by simulating the allocation of predetermined maximum possible amounts of all resources. The algorithm then makes a "s-state" check to test for possible deadlock conditions for all other pending activities, before deciding whether allocation should be allowed to continue.
The Bankers Algorithm needs three things to work:
- How much of each resource each process could request
- How much of each resource each process is allocated
- How much of each resource the system is currently holding The algorithm only allocates resources to processes after it determines whether the processes are in a safe state. Each process releases its resouces only after the process is complete.
We also implemented a scheduling program in traditional FIFO/LIFO in a similar fashion. We used a number of test cases to compare the running time and pass/fail rate of the FIFO, LIFO and Banker's Algorithm.
This implementation will determine if certain state is safe or unsafe (may cause deadlock)
If the given state is safe, which means its all subsequent states are safe, the program will give the safe sequence of the processes and the corresponding available resources in OS
The program also have a timer to measure its performance with different test cases
Result:
| tests | Bankers | LIFO | FIFO |
|---|---|---|---|
| test1 | fail | fail | fail |
| test2 | pass | pass | pass |
| test3 | fail | fail | fail |
| test4 | pass | fail | fail |
| test5 | pass | fail | fail |
| test6 | pass | fail | fail |
| test7 | fail | fail | fail |
| test8 | pass | pass | fail |
| test9 | pass | pass | fail |
| test10 | pass | fail | fail |
| Overall | 70% | 30% | 10% |
Running time (in micro second):
| tests | Bankers | LIFO | FIFO |
|---|---|---|---|
| test1 | 179 | 166 | 207 |
| test2 | 171 | 182 | 145 |
| test3 | 265 | 228 | 217 |
| test4 | 206 | 339 | 334 |
| test5 | 677 | 583 | 548 |
| test6 | 317 | 290 | 289 |
| test7 | 369 | 333 | 311 |
| test8 | 615 | 536 | 1117 |
| test9 | 601 | 146 | 167 |
| test10 | 454 | 338 | 856 |
Bankers algorithm is able to avoid every possible unsafe status. Therefore, it's easy to conclude that for all the test cases Bankers Algorithm fails, both LIFO and FIFO fail as well. These results exemplify the efficiency of Bankers Algorithm in terms of resource management comparing to simple LIFO/FIFO. Bankers Algorithm is more robust, as it succeeds in more test cases than both LIFO and FIFO.
When comparing the algorithms in terms of running time, Bankers tended to be slightly slower than than its LIFO/FIFO counterparts, most likely due to its more complicated structure. FIFO generally performed the fastest, although this may be due to the high failure rate.
Through the tests, there are a few advantages and disadvantages of the Bankers algorithm implementation shown. One of the disadvantages is how much information is required by the algorithm in order to schedule the processes. In certain systems, this information may not be available.
The test cases are provided in the subdirectory "samples"
The test cases have the following format:
P: Number of Processes
R: Number of Resources
int[R]: Available Resources in OS
int[P][R]: Matrix of Maximum Request Resources
int[P][R]: Matrix of Allocated Resources
Sample Test Case:
5
4
8 5 9 7
3 2 1 4
0 2 5 2
5 1 0 5
1 5 3 0
3 0 3 3
2 0 1 1
0 1 2 1
4 0 0 3
0 2 1 0
1 0 3 0
In the sample test case:
- There are 5 processes awaiting resource allocation: I, II, III, IV, and V
- There are 4 types of resources: A, B, C, and D
- For each type of resource, the operating system has
- 8 instances of resource A
- 5 instances of resource B
- 9 instances of resource C
- 7 instances of resource D
- The matrix of maximum requests illustrates the maximum number of resources requested by each of the five processes
- Process I can request a maximum of 3 A-type resources from the system
- The matric of allocated resources illustrates the current state and what resources are currently allocated to the processes
- Process II currently has 2 of C-type resources running
Go Programming Language:
Go was used to implement the scheduling algorithms. Similarly to Rust, Go supports concurrency at the language level and is still in development. The syntax of the language minimizes the length of code, and was similar to other familiar languages.